JPRS ID: 10387 USSR REPORT CYBERNETICS, COMPUTERS AND AUTOMATION TECHNOLOGY

• Document Type: CREST
• Collection: JPRS
• Document Number (FOIA) /ESDN (CREST): CIA-RDP82-00850R000500040033-6
• Release Decision: RIF
• Original Classification: U
• Document Page Count: 127
• Document Creation Date: November 1, 2016
• Sequence Number: 33
• Content Type: REPORTS

APPROVED FOR RELEASE: 2447/02/09: CIA-RDP82-00850R000500444433-6
' FOR UFFICIAL USE ONLY
JPRS L/ 10387 ~ ~
~ 15 March 1982
USSR ~Re~ort
. ~ . ~
- CYBERNETICS, COMPUTERS AND ~ ~
~ AUTOMATION TECHNOLOGY.
~ (FOUO 3/82)~
F~IS FOREIGN BROADCA~T INFORMATION SERVICE
~ .
- ~ , FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
- NOTE
JPRS publications contain information primarily from foreign
newspapers, periodicals and books, but also from news agency
transmissions and broadcasts. Materials from foreign-language
sources are translated; those from English-language sources
are transcribed or reprinte3, with the original phrasing and
other characteristics retained.
Headlines, editorial reports, and materiaY enclosed in brackets
are supplied by JPRS. Processing indicators such as [Text]
or [Excerpt] in the first line of each item, or following the
last line of a brief, indicate how the originai information was
processed. Where no processing indicator is given, the infor-
mation was summarized or extracted.
Unfamiliar names rendered phonetically or transliterated are
- enclosed in parentheses. Words or names preceded by a ques-
tion mark and enclosed in parentheses were not clear in the
original but have been supplied as agpropriate in context.
Other unattributed parenthetical notes within the body of an
item originate with the source. Times within ~tems are as
given by source.
The contents of this publication in no way represent the poli-
cies, views or at.titudes of the U.S. Government.
COPYRIGHT LAWS AND REGULATIONS GOVERNING OWIVERSHIP UF
- MATERIALS REPRODUCED HEREIN REQUIRE THAT DISSEMINATION
OF TFIIS PiJBLICATION BE RESTRICTED FOR OFFICTAL USE ONLY.
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED F~R RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
FOR OFFICIAL USE ONLY
JPRS L/1~3~7
15 March 1982
USSR REPORT ~ ~ �
CYBERNETIC3, COMPUTERS AND AUTOMATION TECHNOLOGY -
(~'OUO 3/8 2 ~ .
~ ~
SELECTED ITEMS ~ROM ~CO'MPUTE'~ TECHNOLOGY
OF THE SOCIALIST COUNTRIES', No, 9, 1981
CONTENTS
- Computer Technology of the Socialist Countri~s .......o.......,......... 1
Technical and Economic Justificatxcn in Selecting Charact~xist;:cs
of Parametric Series of Computers 9
Concept of YsS-1060 Computer Design and De~velopment....e..........~...... 15
Unified Secondary Pow~r SuppliQS for 5eries-2 Unified System Compu'cers.. 22
~,fi$ Preparation Unit Development in Bulgaria 31
YeS-101t5 Computer, Characteristic Features 35
Videoton Famil~~ of Intelligent Alphanumeric Terminals 43
YeS-2335 Array Processor for Ye5-1035 Computer S~rstem 1~9
S~stem for Design of Application Program Pacicages for Calendar-
3ub3ect Planning and Management for Unified S~rstem Computers 52
Digital Simulation of Cantinuous processes on Unified S~rstem Computers... 59
Problem-Oriented Language for S~rstem of Computer-Aided nesign of
Manufacturing Processes of Mechanical Machining of Parts 66
Experience in Use'of Time Sharing System for Program Development 71~
SZPAK-7? Integrated Programming System for Complea Automation 80
- a- [III - USbR - 21C S&T FOUO]
FI~R OF$I~IAL U5E ONtY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
r'UK UFFICIAL USE ONLY
DOS RT Real-Time Operating S`,ystem ................................A....... 87
Softwaxe for Corrnnunication Between Computer System and Specialized
Processors for Arra,y Processing 91
Automated S~stem for Mass Preventive Treatment of Population (SAMPO)...... 96
Data Acquisition and Processing S~stem for USSft Gas Industry 1(?2
Acquisition of Data on Computer Haxdw~re Operation in German Democratic
ftepublic 107
New Small Computer Hardware From German Democratic Republic 113
MPL/600 Algorithmic Language 119
-b-
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500044433-6
FOR OFFiCIAL USE ONLY
COMPUTER TECHNOLOGY OF THE SOCIALIST COUNTRIES
Moacow VYCHISLITEL'NAYA TEKHNIKA SOTSIALISTICHESKIKH STRAN in Ruseian No 9, 1'~81
(signed to prese 21 May 81) pF 2-4, 172-1.73, 176=179 ~ ~
[Annotation, foreword, table of ~contents and abetracts from book "Computer Technolo-
gy of the Socialiet Countries", edited by M. Ye. Raykovakiy, g collectiion of arti-
cles, Izdatel~stvo "Finansy i statistika", 16,Q00 copies, 184 pagea]
~[Text3 Thie international collection deals with problems of research, dev.lopment,
~'~lication and operation of computer facilities developed in accordance with the
~e~'~ement on cooperation in coinputsr technology between the NRB [People's Republic
~D'~ Bulgaria], the VNR [Htingarian People's Republic~, the GDR [German Demoaratic
Republic], the PNR [Polish People's Republic], the Republic of Cuba, the SRR [Socia-
list Republic of Romania], the USSR and the ChSSR [Czechoslovak Socialist Republic].
In the collection, special attention is paid to.prob].ems ef computer applicationa
in ASU [automated management aystems] and computer-aided design syeteme.
The~colle~tion is addreseed to workers engaged in development and use of t~?e
facilfties in the Unif ied Syatem of Computers in various se�~ors of the national
economy.
Foreword
Since 1979, YeS computer usera have been recei'ving computers in the aecond phase
of the Unified System. Industry has now completed fully the transition to produc-
' tion of these new machinea. Tiieir functional capabilities are raising the techni-
cal and economie effect from computer applications in the most varied spherea of
the national economy of the socialist countriea. This ie facilitated alsa by the
- new operating systems that permit implementation of the time-sharing mode in ahared-
use systema, virtual diatribution of machine�~reaourcea, real-time control of .
" objecta, etc. The cagabilitie~ of the new machinea and the systems based on ~hem
, have been expanded through an increase in main etbrage capacity, use of 100M-byte
diak etorage units, and use of new teleproceseing facilitiea and specialized pro-
. c.essars.
The thematic direction of a numbet. of articles i.n this collection ie the develop=
ment of the Unified System of Gomputere. Attother part contains articlea dealing
with comLuter-aided design [CAD] systems (SAPR) based on using the Unified System
~ and System of Slmall Computersp CAD systems ai~e one of the moat effective
- 1
FOR OFFICIAL USE ~ONLY ~
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
ruK ur~r~~~taL USE ONLY
applications of computera. CAD permite reducing the time for design of complex
objects two to three-fold and raises coneiderably their technical and economic
parameters. Articles on the probleme of improving programming are included in
this collection.
The remaining articlea in the collection pertain to problema of application and
operation of the YeS and SM computer hardware and aoftware. Articles on new
hardware and eoftware are in the last section.
The editorial bo~rd for the collection: M. Ye. Rakovakiy, chief.editor
(Coordination Center, MPK [Intergovernmental Commission]), N. V. Gorshkov (Council
on Complex Maintenance of the Ye8 EVM), Kh. Choppe (GDR), Ye. N. Mel'nikova,
responeible secretary (Coor~=ination Center, MPK), B. N. Naumov (Council of Chief
Designers of the SM EVM), L. Nemet (VNR [Hungarian People's Republic]), P. Popov
(NRB [People's Republic of Bulgaria]), V. V. Przhiyalkovskiy (Council of Chief
Designers of the YeS EVM), Yu. P. Seliva~ov, editor-in-chief (USSR), A. Ye.
Fateyev (USSR) and N. I. Cheshenko (Council on Application of SVT [Computer
Technology]). ~
Contents Page
Foreword
' I. International Cooperation Between the Socialiat Countries in Computer
Technology
Selivanov, Yu. P. and Novitskas, Yu. N.. Task of Technical and Economic
. Justification in Selecting Characteriatics of Parametric Seriea of
Computers 5
II. Computer Hardware
Antonov, V. S.; Orayevekiy, K. S.; and Avtonomov, B. B. Concept of the
YeS-1060 Computer Aesign and Development 12
Fayzulayev, B. N.; Batyukov, Ye. I.; and Mkrtchyan, Zh. A.
Unif ~ed Secondary Power Supplfea for "Series-2" Unif ied System
= Computera. 19
- Topalov, T. A. Development of Data Preparation Unite in the People~a
Republic of Bulgaria . 28
Kuchukyan, A. T.; Sarkisyan, T. Ye; and Mkrtumyan, I. B.
Characteristic Features of the YeS-1~45 Computer 32
Nylas, L.; Hudoba, G.; and Bornemisaza, J., "Videoton" Family of
IntE'ligent Alphanumeric Terminals 41
- Nikolov, G. P.; Lazarov, V. D.; Daskalov, P. P.; . Ivanova, Y. V.; and
Kirov, K. D. YeS-2335 Array Proceseor for the YeS-1035 Computer Syatem 47
III. Computer Software
Matulis, V. A. and Chaplinskas, A. A. Syatem for Design of Application
Program Packages for Calendar-Subject Planning and Macsagement for YeS
Computers 51
Yazh~mbek, Ya. Digital Simulation of Continuous Procesaes on YeS
Computers 58
Tsvetkov, V. D.; Tolkachev, A. A.; and Cher, I. Problem-Oriented Language
for CAD Syatem for Manufacturing Procesaes of Mechanical Machining of
Parta 65
. 2
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
FOR OFFtClAL USE ONLY
Brusdeylins, G. Experience in. Use of Time-Sharing System for
Program Development . 73
Aderek, A. SZPAK-77 Integrated Programming System for Complex Automation 81
Shuba, A. DOS RT Iteal-Time Operating System 86
Auerbach, D. Systematic Development of Projects and Programs for
Data Processing 90
Ivanova, Y. and Atanasov, T. .Sof tware for Communication Between Computer
System and Specialized Proceaeors for Array Proceasing 100
Botev, Kh. Ye. Universal Description Language for CAD Systeme 10.5
- Schubert, D.; Boelke, V.; and Ettrig, X. Ttte ALBI-1 Algorithm Storage
and Retrieval System 109
IV. ~?pplicatior. of Computer Facilities
~ Rakovich, A. G. and Samek, A. Concept and Features of Program Package
for CAD of Drilling Devices 117
Pikler, D. and Turai, I. CAD for Dies 124
Astardzhiyan, G.; Bayev, B.; and Raychev, B. ~utomated Syetem for Mase
Preventive Treatment of Population (SAt~O) 127
. Moletz, N. Data A~cquisition and Processing System for USSR Gas Induatry 133
Baumbach, G.-D. Un~.fied Technology for Application Program Development 139
' V. Computer. Operation and Maintenance
Hanisch, E. Acquisition of Data on Computer Hardware Operation in the
GDR 149
VI. Information on New Computer Facilitiea
Leidler, K. New Small Computer Hardware from the GDR 155
Aleksandrov, A.; Yefremova, R.; and Aleksandrova, Zl. The MPL/600
� Algorithmic Language 161
Yezhovich, E. and Vitkova, G. A Concept of Generalized (Standard)
Progra.~n Producte 166
Abstracta
TASK OF TECHNICAL AND ECONOMIC JUSTIFICATION IN SELECTING CHARACTERISTICS OF
PARAMETRIC SERIES OF COMPUTERS
[Abstract of article by Selivanov, Yu. and Novitskas, Yu.]
[Text] Basic features of a family of computers are conaidered as a parametric
series of machines performing data procesaing in various applications. Technical
and economic criteria for optimization of characteristics of parametric series of
computera and ways of so~.ving optimization problems are presented.
3
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
FOR OFFICIAL USE ONLY
CONCEPT OF YES-1060 COMPUTER DESIGN AND DEVELOPMENT
[Abstract of article by Antonov, V; Oraye~skiy, K. and Avtonomov, B.]
[Text] Stage-by-stage development of the YeS-1060 computer, design principles and
characteristics of central units are described. Proceasor atructure, providing high
efficiency in d~.ta processing, and the structure and characteristice of the YeS-4001
universal channel are presented. A properly aelected direction for atage-by-etage
development of a model, that enablea developing a computer with high technical and
economic parametere, is shown.
UNIFTED SECONDARY POWER SUPPLIES FOR SERIES-2 UNIFIED SYSTEM COMPUTERS
[Abstract of article by Fayzulayev, B. N.; Batyukov, Ye. I. and I~Ilcrtchyan, Zh. A.]
- [Text] Basic specifications for computer secondary power supplies are cited. Modu-
lar design of functionally complete power supplies ie justified. Basic technical
and economic featurea of the power aupplies are given. Deaign featurea for power
supply systems for hardware in second phase of YeS EVM and problems of uaing second-
ary power supplies with input without transformers are considered. Specific charac-
teriatica of power supply ayetema for the f irst and second phasea of the YeS EVM
are compared. ~
DEVELOPMENT OF DATA PREPARATION UNITS IN BULGARIA
[Abstract of article by Topalov, T. A.] ~
[Text] Basic functional featurea and technical data of key-to-tape and key-to-disk
units ars considered. Information of the YeS-9002 and YeS-9003 key-to-tape units
- and key~*_~-floppy disk unita developed~and manufactured in Bulgaria ie given.
CHARACTERISTIC FEATURES OF THE YES-1045 COMPUTER
' [Abstract of article by Kuchukyan, A. T.; Sarkisyan, T. Ye.; and Mkrtumyan, I. B.]
[Text] Basic parametera of the YeS-1045 computer, set of atandard and optional fa-
cilities and ways of obtaining new configurations are analyzed. The etructure of
the CPU and characteristics of its componente are given. A highly developed check-
. ing and diagnostic syatem, providing higher reliab~ility and repairability for the
machine, ~is described. Syatem capabilities of the model and organization of~the ~
array procesaor and additional capabilities that result when it is connected are
discussed.
VIDEOTON FAMILY OF INTELLIGENT ALPHANUMERIC TERMINALS
[Abstract of article by Nylas, L.; Hudoba, G. and Bornemiesza, J.]
_ [Text] General trenda in terminal development are discussed. Methods of developing
terminals with alphanumeric diaplays at the Videoton enterprise are shown. Func-
- tional capabilities of new video terminals with microprogram control, based on
microproceaeors, for the YeS and SM computere are described. Technicai character-
istics for the SM-7219, YeS-7168N, SM-7401 and UT-20 units are.given.
,
~
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/49: CIA-RDP82-00850R040500040033-6
FOR OFF[CIAL USE ONLY
YES-2335 ARRAY PRCICESSOR FOR THE YES-�1035 COMPUTER SYSTEM.
[Abstract of article by Nikolov, G. P.; Lazarov, V. D.; Daskalov, P. P.; Ivanova,
Y. V.; and Kirov, K. D.] .
[Text~ The YeS-2335,.array processor functional capabilities, general atructure and
methods of connection to the computer are described. It permita a significant in-
crease in computer throughput for special taeks.
SYSTEM FOR DESIGN OF APPLICATION PROGRAM PACKAGES FOR CALENDAR-SUBJEC~ PLANNING AND
MANAGEMENT FOR YES COMPUTERS
[Abstract of article by Matulis, V. A. and Chaplinskas, A. A.]
[Text] Basic system characteristics are given; package desigr, technology is des-
cribed. Given as an example are program packages_for five-y~ax planning of program
development in the machine-building sector, and for quarterly and annual plannin~ of
program development for enterprises uaing the system described.
DIGITAL SIMULATION OF CONTINUOUS PROCESS ON YES COMPUTERS
[Abstract of article by Yazh~mbek, Ya.]
- [,Tpxt] The DIANA problem-~rient~d programming language, used to deacribe a system
~o~be aimulated in digital form, ie presented. The language permits solving prob-
L~1s.~written in the form of differential equations or an analog block-diagram. The
basic functions of the language, all its functional atatenaenta and the structure of
a DIANA program are given. Advantages of this language compared to other similar
ones, CSMP for example, are conaidered with regard to the need of.solving ~roblems
on the small YeS models.
~ PROBLEM-ORIENTED LANGUAGE FOR CAD SYSTEM~FOR MANIJFACTURING PROCESSES OF MECHANICAL
MACHINING OF PARTS
[Abstract of article by Tavetkov, V. D.; Tolkachev, :A. A.; and Cher, I.]
[Text] The language ia intended for deeign of informational models of pares to be
~ proceased in systems for CAD of manufacturing processe~s of inechanical machi.aing af
parts. Facilities and rul~es for formalized description of atructural elements and
machine parta as well as the dimension and preciaion linka between these elements
are given. Zhe language may be uaed in CAD $ysteme for industrial procesaes and to
prepare control programs for machine toola with numeric control.
EXPERIENCE IN USE OF TIME-SHARING SYSTEM FOR PROGRAM DEVELOPMENT
[Abatract of article by Brusdeyline, G.]
[Text] A year's experience in uaing facilities of a time-aharing syatem to improve
development of problem-oriented software is deacr.ibed. The individual atagea of
program compilation in the time-sharing mode are discusse~l in detail; technical and
economic effects are evaluated.
5
FOR .OFFICIAL ~JSE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/49: CIA-RDP82-00850R040500040033-6
FOR OFFICIAL USE ONLY
SZPAK-77 INTEGRATED PROGRAMMING SYSTEM FOR COMPLEX AUTOMATION
[Abstract of article by Aderek, A.]
[Text] The baeic elements of the SZPAK-77 syatem are presented. It is a form sys-
tem of programming. Its language allows convenient writing of data acquisition and
preprocessing algorithms. Proceas variable processing algorithms can be extended
by using auxiliary programa in the extended RT-FORTRAN-IV-E language. The SZPAK-77
system has an interactive language for communication with the operator to input
changes into the system, expand it, initiate standard reports, etc.
DOS RT REAL-TIME OPERATING SYSTEM
- [Abatract of article by Shuba, A.]
[Text] The featurea, purpose and degree of complexity of the DOS RT for the SM-3
and SM-4 minicomputere are described. The system offers the capability of creating
an environment of several concurrently operating progrem proceases that interact
with each other and with interactive operators and external processors.
SYSTEMATIC DEVELOPMENT OF PROJECTS AND PROGRAMS FOR DATA PROCESSING
[Abstract of article by Auerbach, D.]
[Text] A unified approach for developi:ng data processing projecta at the atages ur
design, implementation and teeting is suggested and based on modern programming
technology. The basis of this approach ie the systematic decomposition of the task
~ into a hierarchy of aubtasks with well-defined rules of subordination for transfer
of control, organization of the data atream and error processing. Brief information
is given on the most important principles at each stage of development, implementa-
tion and teating. Current ahortcomings in the method and ways of updating it are.
shown. ~
SOFTWARE FOR COMMUNICATION BETWEEN COMPUTER SYSTEM AND SPECIALIZ~D PROCESSORS FOR
ARRAY PROCESSING
LAbstract of article by Ivanova, Y. and Atanaeov, T.]
CText] Features of software aupporting methode of accesa to array processors are
deacribed. Ttao cases are considered: the YeS-2335 array~processor without its own
_ storage and the array procesaor with its own storage. The aet of matrix operations
eupported is given. . .
UNIVERSAL DESCRIPTION LANGUAGE FOR CAD SYSTEMS
[Abstract of article by,Botev, Kh. Ye.]
[Text] The language, developed for writing, inputting and updating data in a data
file for a CAD syetem, features flexibility, universality and eimplicity. A trans-
lator is provided. In each specific case, the user himself def inea the atructure
and contents of the input data stream. Data on the file structure and rulea~for
filling fielda are etored in a permanent eystem file. The facilitiea developed may
be used in any aystem with filea of the type coneidered.
~ ~ FOR OFFIC ~ u$E ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2407/02/09: CIA-RDP82-00850R000500440033-6
FOR OFFICI,AL USE ONLY
AI.BI-1 ALGORITHM STORA~E AND RETRIEVAL SYSTEM
(Abstract of article by Schubert, D.; Boelke, V.; and Ettrig, X.]
[Text] Thie system is based on the AIDOS programming system, which establishes a
uttified technique and criteria for algorithm description. The description laegtlgge
and elements of the retrieval language are ~iscussed.
CONCEPT AND FEATURES OF PROGRAM PACKAGE FOR CAD OF DRiLLING DEVICES
[Abatract of article by Rakov~.ch, A. G. and Samek, A.]
[Text] Considered are the principles for design of a package and its tnain compon-
ents with emphasis on describing the library of prc~gram modules for syntheais and
documentaing designs. Inforniatian is given on hardwRre for implementation of the
- device design programs; a brief economic evaluation of the use of the package is
~ given; information on prospects for its development and improvement ia given.
CAD OF DIES
[Abstract of article by Pikler, D. and Turai, I.]
[Text] Generalized is the experience of developing a system for computer-aided de-
'~~:gn of dies, in which an attempt is made to solve the problem of limited knowledge
~q mechanical engineers of computera and the limited capabilitiea for writing algo-
"t~ithms for design procedures.
AUTOMATED SYSTEM FOR ME:SS PREVENTNE TREATMENT OF POPULATION (SAt~O)
[Abstract of article by Astardzhiyan, G.; Bayev, B.; and Raychev; B.]
[Text] This system for preventive treatment of the population is an appli~ation
program package to be used in real time under control of the SUIP program system,
based-on YeSTEL teleprocessing hardware. . Hardware, package functional features and
data bank organization are conaidered.
ACQUISITION AND PROCESSING SYSTEM FOR USSR GAS INDUSTRY DATA
[Abstract of article by Moletz, N.]
[Text] Reaults of apecial software development are given. General review of sys-
tem etructure and ita componenta ie given; areas of effective application are shown.
UNIFIED TECHNOLOGY FOR APPLICATION PROGRAM DEVELOPMENT
[Abstract of article by Baumbach, G.-D.]
- [Text] Described ia technological process of developing application programs. The
technology ia suitable for developing large program syatems:characterized by diverse
functions, high complexity and dynamic development of area of application. Process
begins with formulation of req~iitementa for the program and is continued with design
of the h~erarchy of functions, data and documentation.
7
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2407/02109: CIA-RDP82-00854R000500040033-6
FOR OFFICIAL USE ONLY
ACQUISITION OF DATA ON COMPUTER H~WARE OPERATION IN GDR
[Abstract of article by Haniach, E.]
['Text] Experience gained in use of system for acquisition and analyais of data on
computer and device reliability in GnR is described. The system effects continuatid
monitoring of the entire computer park and allows taking timely etepe to maintain
high reliability of manufactured hardware.
NEW SMALL COMPUTER HARDWARE FROM GDR
[Abstract of article by Leidler, K.]
[Text] Featurea, operational capabilities and brief information on aoftware for
two microcomputera, a printer and a mark reader in the SM family are presented.
MPL/600 ALGORITHMIC LANGUAGE
[Abstract of article by Aleksandrov, A; Yefremova, R.; and Aleksandrova, Z1.]
[Text] The MPL/600 algorithmic language oriented to the SM-600 microproceasor fami-
ly and the MPL/600 crosa-compiler for the Unified System of Computers are considered.
Information on required computer configuration and average compilation time is gi~-~n.
Described are language capabilities and features, compiler operation, language s;
tax and rules for writing individual atatemente. Difference between MPL/600 and
PL/1 is pointed out. ~
CONCEPT OF GENERALIZED (STANDARD} PROGRAM PRODUCTS
[Abstract of article by Yezhovich, E. and Vitkova, G.]
~ [Text] Deacribed is the ADAPTOR program system, a formal meana for creating and
using flexible program products: ~tandard elements of aoftware for ASU~s of varying
sca2e and generality. The aystem belongs to the class.of macroprocessors which do
not depend on the type of source language.
COPYRIGHT: Izdatel'stvo "Fin~nsy i etatistika", 1981
8545
CSO: ~ 8144/0138
8
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED F~R RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
FOR OFFICIAL USE ONLY
TECHNICAL AND ECONOMIC JUSTIFICATION IN SELECTING CHARACTERISTICS OF PARAMETRIC
SERIES OF COMPUTERS ~
,
Moscow VYCHISLITEL'NAYA TEKHNIKA SOTSIALISTICHESKIKH STItAN in Ruasian No 9, 1981
_ (signed to presa 21 May 81) pp 5-11 .
[Article by Yu. P. Selivanov, engineer (USSR), and Yu. M. Novitekas, candidate of
economic science (USSR)] .
'[.'fiext] The principlea, developed in machine buildfng, for creating.parametric
,se~ies of machinea are also applicable ta computer science ae the field concerned
i+~tt'~ the design, production and operation of da'ta proceasing hardware. The use of �
these principles may be of coneiderable helg in developing general approaches to
- the development of families of computera and fn quantitative and qualitative
- evaluation of the ch~racteristica of such a.fart~ily.
In accordance with [1], as applied to computere, a parametric seriea should be
understood as the aggregate of ite elements--hardware, software and data compatible
type sizes of computers for an operational puZpose, 'i.e. intended f~* uae in a
specific macro area, hav.ing unif isd architecture, having limited interchangeability,
differing from each other by the numeric value of the main parameter, throughput,
and intended for meeting the given needs of society for processing of data.
As apparent from the above definition of a F~arainetric seriea, throughput is usually
accepted as the main computer parameter. On the one hand, thia parameter is closely
tied to all the other computer characteristics and of ten prescribes them, and on the .
other hand, it moat fully reflects both the user propertiea of these machines and
the cost for their production and operation.
In our view; based on the general theory for design of parametric seriea of machines,
in the most general serlse, throughput should be thought of as an indicator that
takea into account not only the effective rate of the aggregate of functional unita,
but also the labor productivity of the pereonnel preparing the process of data pro-
ceseing and participating in it. This meane f~iat cofnputer throughput muat be
spoken of as a man-machine system, which it basically is. Hence, it followe that
the problem of increasing computer throughput must include the very.important
partial problem of reducing the ahare of coet of direct labor in the overall aggie-
gate of direct and indirect labor needed to process a unit of information in solving
one type problem. The way to solve this partial probl~em is to transfer ~to hardware
- and system software facilities for implementation in the automatic mode of those
functions for solving the problem (development of the algoritt~m and program,
� 9 ~
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2407/42/09: CIA-RDP82-40850R000500440033-6
FOR OFFICIAL USE ONLY
preparation of data, providing for an efficient mode of procesaing, operation of
hardware, etc.) that have been previously performed manually in parctice. This
- leads not only to a decrease in total inputs of time needed to process a unit of
information, but also to an increase in these inputs for the share of machine time
at the expense of the manual, which is a mandatory prerequisite for further autottta-
tion of all the procesaes in data processing in the national economy based on
qualitatively new computer hardware.
This methodological approach to the evaluation of the throughput of a parametric
series of computers permita determin~ng the technical and economic effectiveness of
the application of this parametric series in the entire macro area for which the
series is designed. Also, this approach does not preclude the usual methods of
evaluating throughput that permit determining the physical parametera af computers
in solving problems of a apecific type from the given macro area.
A major property that computere forming a paremetric series muet have is inter-
- changeability. The classification by features of interchangeability of t~pe sizes
of a parametric series of machinea and the characteristics of each typE are.given
in [2]. The problem of interchangeability of computers as type sizea of a para-
metric seriea has not yet been adequately studied and requires aeparate considera-
tion. By limited interchanegability, which computers in this aeriea muat have.
according to the definition of a parametric aeries given above, is meant the capa
bility of replacing one type size of the aeries by at least onP other, which, how-
ever, entails certai~ economic lossea, for example, the substituting computer
solves the same problems at the same qualitative level as that being replaced, but
it costs more to do so.
A prerequisite and mandatory condition for interchangeability of computers forming
a parametric aeries is their hardware, aoftware and data compatibility.
' Hardware compatibility of computers in a parametric series provides the capability
_ of building them on the modular principle, i.e. by assembly of all computera in a
parametric eeries from a unified set of fuactfonal devices, aggregatea that are a
set of structurally and functionally unified standard unita and assemblies connected
to each other by unified links. A relativelq emall number of unified units and
assemblies is needed to obtain a large number of modificationa of functional devices
built on the modular principle. This approach reduces hardwar~ development and~
manufacturing time and simplifies and reducea the coat of operating it.. Unification
of interfaces creates the capability of improving any of the devices independently
of the others with subsequent inclusion of them into the configurati.on of a machine
already in operation. This enables the development of the techni~al capabilities
of the individual members of ~he parametric series, preventing obsolescence and
exCending thereby the life cycle of the aeries.
The modu~.ar principle permits developing the technical capabilitfea of an existing
computer not only in depth, by improving the individual functional unita of the ma-
chine, but also in width, by controlling the machine configuration. Using this
principle of designing hardware combined with eneuring aof tware and data comp+atibil-
ity and the uae of apecial additional facilitiea permits forming complexes of com-
puters, creating structures with a qualitatively new level of system organization
and qualitatively new technical and economic parameters.
10
- FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/49: CIA-RDP82-00850R040500040033-6
,
~OR OFFICIAL USE ONLY '
Hardware, aoftware and data compatibility of computers in a type eize series ie
based on the unity of their architecture which hae a modular etructure and the
development of v?hich therefore can be effected without losa of software compati-
bility of the machfnes with computers previously manufactured.
A brief consideration of the characteristics of the propertiea of computers as
type sizes of a paraanetric aeries permita formulating the requirements imposed on
them from this viewpoint. They must be hardware, software and data compatible,
have the properry of interchangeability, have flexible configuration and lend them-
selves to complexing. Theae requirements, in particular, have been fully met by
the second phase of the Unified Syetem of Computers. In the first phase, the;re-
quirement of inetrchangeability wae not fully realized since the YeS-1010 and
YeS-1021 computers do not have the property of interchangeability with the other
. machines,
The features peculiar to a parametric series of computere and which characterize it
as a type size series of machines are multidimensionality and multilevelnesa.
Multidimensionality of a parametric series of computers is expreased in that the
machines forming it differ from each other not only by the numeric values of the
main parameter, but also by the numeric values of other parametere, baeic and
s~acondary, for example, by tihe main storage capacity, channel throughput, floor
s~ace occupied, etc. In some areas of application, theae additional parameters
may:turn out to be the main ones. For example, for information syatems with large
external storage, channel throughput may be the.decisive parameter in choosing a
machine.
The multilevel nature of a parametric series of computer.s is manifeated in the
fa~ct that not only the comput.ere, but also the tunctional units forming them and
th~p modules of the~lower levels, including the integrated circuita, are elements
of specific parametric aeriea. In other worda, each computer, being an element of
a l~arametric series of these machines, consista of devicee, units and elements,
etc:. that in turn are elements of a corresponding partial series. The multilevel
prc,~perty is especially characteristic of the second phase of the Unified System of
Com~puters. In the first phase, only central units formed paremetric series. One
could single out groups of peripherals that did not form a parametric series for
� the main parameter, for example, aeveral typea of diak storage units wfth identical
cap,acity of 7.25M bytes.
A m~jor characteristic of a parametric asries of computers, just as of any other
_ part~metric aeries, ie its denaity, i.e. number of type sizes. Parametric series of
com~utere L~r,.,,,Sle of performing the same amount of computational operations at the
same qualitative leveZ may cansist of a diverse number of type sizea (differ in
denaity), and aeries with an identical ncunber of type sizes (equal in dens3.ty) may
consist of type eizes diverse in the numeric ~�aiue of the main parAmeter.
Thus, a certairi amount of work on data processing may be performed with the aid of
different psremetric series of comDUtere. One cannot compare computere in a para-
metric series with separate computers that do not form this seriea, since the para-
meters of~ the computers forming the parametric series are selected and optimized on
the bffisis of the tasics set for the entire series. The.uaer coat of data procesaing
in al~l these alternatives ia identical aince they permit performance of the same
~ amount of work at the same qualitative level, but the labor cost is not equivalent
11
FOR OFF[CIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007142/09: CIA-RDP82-40854R040500040033-6
FOR OFFICIAL USE ONLY
~ince the performance of the same amount of work requires different eocially
needed inputa of labor. In other words, inputs for develo~ment, production, com-
missioning and operation of computers depend on how the series is constructed (on
the number of type sizes and denaity of their arrangement in~its individual inter-
vals, defined by the numeric value of throughput peculiar to the models in this
series).
Consequently, the problem of chooaing the optimum parametric series of computera
from all possible alternativea hae to be solved. To solve this problem, it ie ex-
pedient to use the method of comparative economic effective~ess, the calaculation
of indicatore of which is based on monetary measurement of inputs. The inputs
listed above asaociated with the development of parametric aeries of computers are
synthesized in two monetary indicatore: capital inveatment and operating expensea.
Capital investment associated with development of the i-th alternative of a para-
metric series r.^.~y be defined by the formula.
m
Kpg ~ - ~~1 ' ~COM'~' ~ 1)
I-
where KCOM i3 is the capital investment needed for development, manufacture and
commissioning of a computer of the j-th model of the ~'_-th parametric aeries,
rubles/computer. .
, The value of KCOM ij for each j-th model of computere in the series has to be
determined with regard to the wholesale price for the minimum configuration of the
model; fnputs for design of the parametric series, including inputa for development
of the sof tware ayatem per model (if for some reason these costa are not in~luded
in the wholeeale price for the machine); inputs for tranaportation of the computer
to the installation site, assembly, adjustment and startup of it; and costs of
machine floor space occupied (including space needed for aervicing and repair).
Expenses for operation of each i-th alternative of the design of a parametric
series of computers can be calculated as
~
ps `~ON~~~' ~2~
where ECOM ij is the annual expense for operating a computer of the ,j-th model of
the i-th parametric seriea, rubles/year. ~
computer
The value of ECOM ij muat include: baeic and supplementary ealaries for personnel
_ servicing the type configuration of the machine; amortization of the computer;
coat of servicea performed by apecialized enterprisea for complex centralized
maintenance of computer hardware per machine; coet of routine computer repair per-
formed by maintenance personnel; coet of power consumed; coat of auxiliary materi-
als (magnetic tapes, perforated tapea, paper, etc.); and amortization of machine
floor space occupied.
12
� FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
- FOR OFFICIAL USE DNLY
The economically optimal alternative of a parametri~ series of computera is chosen
based on the monetary criterion of comparative effectiveneas, the indicator of unit
adjusted costs for the i-th alternative of a design of the series. This indicator
is determined by the expression
~ m m
- W t } _ ~1 E~~ il + `n~~.+~ ~Ol~ ~I 3
~l11= 9ps` m . ~ )
qCtJM l/
where ~n is the normative f actor of eff ectiveneae of additional capital inves~nent,
1/year;
qps i is the total annual throughput of all models of the computers forming the i-th
parametric series, thoueands of operations/year~
series of computers _
and qCOM ij is the annual throughput of a computer of the j-th mo~el of the i-th
parametric series, thousands of operations/year r type taska/year e~c.~
computer l computer
~Considered economically optimal ie that i-th alternative of the design of the para-
` ,m~;tric series of computers by which ~
W d ~i = min. ~4)
Thus, based on what has been said above and with regard tn [1], an economically op-
timal parametric series of computers should be coneidered a parametric series of
theae machines with auch numeric values of the main parameter, throughput and other
(basic and secondary) parameters with which the apecified natidnal economic needs
for data processing are met with the least unit adjusted costs.
The basic requiremente imposed on an economically optimal parametxic series of com-
puters are: ~
long-term nature~, ensured by including in the series future type sizes of computera,
the need for which cannot be met immediately or rnay arise later;
- optimality of period of manufacture, i.e. an economically optimal term for the
aeries models being in producCion, determined by the technical level and quality of
theae models, rate of technical progress in computer manufactu~e and sectors re-
lated to it, and rate of change in the national economy of the volume and structure
- of information subject to processing on computere of thie type;
technical and economic juatification which includea a technical and economic
analysis of the parameters of computers manufactured by domestic and foreign indus-
try; selectior~ of computer parameters determining the technical level and quality
of these machines; analysis of the actual data and forecasts for the volumea and ~
atructure of information subject to processing on computera of a given type; defi-
nition of the need by the national economy for auch machines; selection of a
rational ar3 Frogressive base design for the computera and the sof tware aystem
(from the viewpoint of performing computatior~s with ttie minimal socially needed
inputs of labor); determining and comparing cost required to implement the different
alternatives of a parametric aeries of ~omputers of a given type, etc.;
optimality of nomenclature.of computer technical parameters prescribed by standards,
and selection based on objective criteria not only of the main, but also of the
baeic parameters of the machines from the set of exiating parametera.
13
' FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500040033-6
" FOR OFFICIAL USE ONLY
In setting out to design a parametric aeriea of computers, first of all one must
establiah ita ~~;era, the moat typical~operating conditione and the demand by the
national economy for the machines of the various type sizes. After this, the al-
ternatives of the parametric series have to be outlined and the alternative that
most fully meeta the requirements of production adn operation of these machine~
has to be chosen on the basis of technical and economic analysis and with regard to
the criterion of optimality (4).
The starting point for development of an economically optimal par,ametric series of
computers ia the definition of demand by the national economy for the gi,ven type of
machines. In turn, computer demand is based on the data volume and structure that
it is economically expedient to procesa by ueing machines of a given type. Great
difficulties arise in analyzing the demand for tasks to be handled by using compu-
ters and the distribution of these tasks by parametere deacribing them. Still,
this is the only kind of analysis that permits defining the demand by the national
economy for computers of specific type sizes with technical parametere optimally
corresponding to parametere af the tasks to be handled. ~
BIBLIOGRAPHY
1. ~~Tipovaya metodika optimizatsii mnogomernykh parametricheskikh ryadov"
[Standard Tectinique for Optimization of Multidimensional Parametric Series],
Moscow, Izdatel'stvo atandardtov, 1975. ~
2. Zvanov, A. V., "Ekonomika ryadov mashin" [Machine Series Economics], M~scow,
Izdatel~stvo standartov, 1975.
COPYRIGHT: Izdatel'stvo "Finansy i statistika", 1981
, 8545
CSO: 8144/0138
11~
FOg O~k'~.CTAL U$E pNLY '
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500044433-6
FOR OFFICIAL USE ONLY
CONCEPT OF YES-1060 COI~~'UTER DESIGN AND DEVELOPMENT
Moscow VYCHISLITEL'NAYA TEKHNIKA SOTSIALISTICHE~KIKH STRAN in Rusaian No 9, 1981
(signed to preas 21 May 81) pp 12-19
[Article by V. S. Antonov, candidate of engineering science (USSR), K. S.
Orayevskiy, engineer (USSR), and H. B. Avtonomov, engineer (USSR)]
[Text~ ~A trend in the evolution of the Unified System of Computers is the improve-
ts~m~, of models previously produced by extending their functional capabilities. Such
d~e~wc~.opment a~aumes maximum preservation of the basic structure and conceptual solu-
t~cae~s adopted in the central units of the computer, the processor, the IO channels
and the main storage unit, and of their interface principles. This permits extend-
ing the term for series production of the computer and at the same time provides the �
capability of developing for the computer newl~y introduced functions, principles for
implementation of them and new versiona of the operating syatema to facilitate their
introduction and ensure continuity in the models being produced. This approach to
computer development as a whole becomes especi~ally important in developing the pro-
ceasor, since introduction of new functions, primarily expanding the instruction
set, introducing new types of interruptions, e~c.. is reflected mainly in the pro-
cessor.
~Now, when the prograa? for further development of the Unified System of Computers has
been formulated, it is useful to generalize the experience of developing the
YeS-1060 computer. ~
In desfg~ning the YeS-1060 computer, the following tasks were solved:
development of a high-throughput computer that implementa the principles of apera-
- tion of the second phase of the Unfied System of Computers;
raising the technical and economic characteristics compared to those of the compu-
ters in the f irst phase of the Unified System;
raising the internal speed of the processor; .
increasing the capacity of.main storage;
raising.operating reliability;
development of checking and diagnoetic facilities; and
introduction of the new technological design and element~ base.
Development of the YeS-1060 computer began while the principle.s of operation for
the second phase of the Unified System of Computezs were still being worked out, but
the basic concepts had already been formulated and adopted, such as the organization�
and facilities for providing virtual storage, expanding the inatruction set and the
15
- FOR OFFICIAL USE ONLY�
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/49: CIA-RDP82-00850R040500040033-6
ruK ~rr~~iwl. U~~. UNLY
system of interruptions. The YeS-1060 computer had to become ,g tranaition model;
therefore, it was oriented to further development. In doing sc+, a path of stage-by-
stage development was chosen, and the computer of the first stage had to become a
series model. Let ~is consider the ways and emthods of solving the basic problems
in designing the YeS-1060.
The main technical and economic characteristic of a computer is the throughput-cost
ratio. Thia characteristic can be improved by either increasing computer through-
put or reducing hardware cost. In turn, throughput can be increased through both
more efficient orgarization of the computing process, raiaing the level of ~ulti-
programming anc reducing unproductive inputs of machine time, and through increasing
the internal speed of the processor,~IO channel throughput and rate of operation of
the external devices. However, increasing processor speed and IO channel throughput
requires additional inputs of hardware, which increases computer cost somewhat.
Increasing computer throughput requires in the first place expanding the capacity of
main storage which enables eff icient use of the computer.
In the YeS-1060 computer, the technical and economic characteristics have been im-
- proved as a result of the optimal coinbination of t~ie methods mentioned. Thus, for
_ exarnple, considering the r,apabilities or t~h element base used, a relatively simple
structure was selected for the processor with overlap of instruction processing ~
� three levels and application of efficient microprogram control, which permitted u.-
taining a throughput of over one million operations per second with comparatively
loW costs for hardware.
A 18rge main storage unit was developed in the YeS-1060 computer in the first stage
Chrough the fncorporation of mrr~ modern design solutions for the ferrite storage
used previously. This enabled quadrupling the atorage capacity in a rack. In the
second stage, integrated circuita were used ae the storage medium.
A major problem in developing a computer is raising operating reliability, which
means faultless operation, automatic recovery of the computing process when system
serviceability is damaged and reducing repair and maintenance time. The checking
and diagnostic facilities are aimed at accomplishing this.
Parity checking facilities incorporated in the YeS-1060 computer cover up to 90 per-
_ cent of the equipment in the main processing assemblies and data transmission paths.
Hamming code checking was also introduced for main storage. This enables detection
of single and double errors and correction of single errora nf data in etorage,
Diagnostic facilities include both sof tware and microdiagnostic teats that together
with the checking facilities permit detection of practically all malfunctions or
f ailures of central unit hardwase, procesaor, storage and channels, and localization
of malfunctions with a precision down to a group of TEZ's [standard exchange cards].
These facilities reduce considerably computer maintenance and repair time and also
provide for operation of recovery facilities. The latter include hardware retry of
' instructions and procedures in the CPU and channels and sof twaie facilities in the
operating system that provide for retry of programs or sectiona of programs. ~
Standard solutions for the Unified System of Comp~sters were adopted ae the basis
for mechanical processing of the YeS-1060 computer. To raise the level of machine
16
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500040033-6
FOR OFFICIAL USE ONLY
manufactureability in series production and imprqve operating qualities, a numFi~r
of new industrial procesaes was intrcduced in machines of thia clasa, such as wire-
wrapping in assembly of panels, assembly of inter-panel connections by flat cable,
etc.
ECI,, the IS500 series of inetgrated circuits, was chosen as the computer element
base. This series includes 26 types of circuits with medium density of about 10
gates per package with typical delay time of 1.5-2 ne per element and average dis-
sipated power of about 250-300 mW per package. Subsequently, the number of types
of integrated circuits reached 47 as a result of introduction of more complex func-
tional elements and development of storage elements. Average number of gates per
package grew to 15-20.
Uaing the IS500 made it posaible to implement an advanced structure for the central
units in the computer with ra~her limited structural bulk and with a relatively
short operating cycle.
- The Processor. In analyzing the capabilities of improving the processor structure,
the conclusion was drawn that units such ae the instruction processing unit and the
arithmetic unit control unit must be the most dynamic and adaptable. Analysis of
~1}~,experience in developing and adjusting the YeS-2060 processor showed that the
~~e~~~est difficulties are encountered in developing the control circuits.
Th~ primary use of microprogram control and the use of loadable storage of micropro-
grams allow for development of the processor, introduction of new functions and
' extension of the instruction set without substa~tial adjustment to the apparatus.
Microprogram control also provides efficient diagnostic facilities that can be de-
- veloped without additional costs for equipment. The expediency of this solution
was confirmed both during adjustment of the pilot model and during development of
new versiona of the opera~ing systems.
The YeS-2060 processor (fig. 1) has three basic units: the storage control (UP)
unit, the central control and instruction processing (TaU) unit and the arithmetic
- unit.(BA). The storage control unit services requests coming from the proceasor
units and IO channels for reference to main storage. The main functional assemblies
in the storage control unit are high-speed buffer storage (BP), IO channel buffer
storage (BK) and dynamic address translation facility (DTA) and the st~rage adapters
(AP). The central control and inetruction procesaing unit fetchea instructions from
main storage, processes them in the overlap mode, generatea operand addresses and
fetches operands from local and main storage. Its functional assemblies include
the buffer registers for instruction pr.ocesaing (RB), the instruction counter
(SchK) and the microprogram control unit (MU). The arithmetic unit performs opera-
tions on operands in the process of executing operations and writes the results to
local and main atorage. The main functional processing asaemblies in the arithme-
tic unit are the parallel adder (SM), the shifter (SD) and the fast multiplier (UM).
The control units for the central. control and instruction processing unit and for
the arithmetic unit are microprogrammed. Contrql signals are generated by the
microprogram control units.. ~
~ The chosen structure of the processor permits solvfng the problem of matching the
relatively long access time from main storage with the short CPU cycle and provides
for efficient overlap of the operation of the IO channels and CPJ with main storage.
17
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007142/09: CIA-RDP82-40854R040500040033-6
FOR OFFICIAL USE ONLY
_ ~ ~ xey:
~3~bJ~.~XR~',.~+,P~e 1. UP [storage control unit]
2. AP [storage adapter]
Y~ 3. BK [IO channel buffer storage]
(1) o ~ s ~ 4� BP (high-speed buffer storage]
~n 5. DTA [dynamic address translation]
- 6. TsU [central control and instruc-
tion processing unit]
_ 7. SchK [inatruction counter~
_ p~~ ~4e~ ar~ 8. RB [buffer regiaters for
~ instruction~processing]
9. MU [microprogram control unit]
- 10. BA [arithmetic unit]
. ~6~ 11. SM [parallel adder]
c~K pQ My 12. SD [ehifter]
13. UM [fast multiplier]
6~
y
~ ~ \ ~ �
Fig. 1. Structural diagram of the YeS-2060 processor .
These problems are solved by incorporation of the mode o~ interleaving of main
atorage (interlacing of addresses), the high-apeed buffer storage and the buf�er
storage fcr the IO channela in the storage.control unit. ~
The specified epeed of the CPU is achieved through, first, overlapped processing of
instructions which permits obtaining results of execution every two stepa of the
processor operation; second, more efficient structure and algorithms for execution
of operations in the arithmetic unit which decreases the average number of stepe in
instruction execution; and third, as a result of incorporation of the high-speed
buffer storage which reduces the effective cycle of main storage threefold.
CPU characteristica are:
principle of control hardware-microprogram
instructions realizabla 183 .
overlap levela 3
throughput, measured on the basis of
statistical mixes for scientific and
technical taska ~ 1,040,000 instructions/second
operating step 160 ns ~
buffer storage: capacity 8K bytes
cycle 160 ns
~ 18 ~
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/49: CIA-RDP82-00850R040500040033-6
, FOR OFFiCIAL USE ONLY
The YeS-4001 Univeraal Multiplexer Channel extende the functional capabilities of
the computer IO system. It unites groups of functionally independent channels in
one device (fig. 2). The byte-multfplexer channel (BTMK) s~rvices low-speed
~$~~o~sor ~y ~
1. BOM [unit of exchange buses]
2. BT1~IIt [byte-multiplexer channel]
3. GK [main channel]
~1~M 4. PIrIIC [local storage for main
channel]
' S. SPK (0, 1, 2, 3) Lselector
subchannels]
6. 1~K [block-multiplexer subchannel]
srMK~2~ ~~?K 6AMK1 ~~MK: BLIrIICO [block-multiplexer channel]
, (7 (11) (12) 8� BSTsP [CPU interface unit]
(g icun . 9. ~BBD [data buffer unit]
~ rx nn~r 10. BSPU [peripheral interface unit]
~ ~9 iaa . 11. BI.IrIItl [block-multiplexer channel]
12. BLI~M2 [block-multiplexer channel]
nK crtK enK crtx ,ynx , bC119
!0. 1 1 J
1~
NMm~vl~+~c aol.�~ra1s
IO interface '
_ Fig. 2. Structural diagram of the YeS-4001 universal multiplexer channel
peripherals in both the multiplexer and burst modes, and medium-speed devices in
the burst mode. The basic functional units of the byte-multiplexer channel are
the main channel unit (GK), four selector subchannels (SPK) and the block-multi-
plexer subchannel (MPK). The main channel unit enables exchange of signals with
the CPU and storage control unit, exchange of data and control words with main
storage,.and storage and modif ication of control data. The main channel exchanges
data with the CPU through the.unit of exchange buses (BOM). The main channel ex-
changes data with the multiplexer subchannel in bytes. Storage and unpacking of
double words is performed in local storage (PIrIIC) for the main channel. Selector
aubchannels have their own data buffera and exchange data in double words with the
main channel.
A block-multiplexer channel services high-speed peripherals. The device is con-
nected through a fast two-byte interface, permitting data transfer in both the one-
byte and two-byte modes. The basic functional units are the CPU interface unit
(BSTsP), the peripheral interface unit (BSPU) and the data buffer unit (BBD). Data
is exchanged with the CPU through the unit of exchange busses in double words over
a bus common to the block-multiplexer channels. ~
The block-multiplexing mode allows a.considerable increase in IO system efficiency
and eff icfency in the use of channel equipment. The universal channel supports: ~
the indirect addressing mode in the channels;
19
~ FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500044433-6
r~uK ur~r~ic:tA~ USE ONLY
the two-byte interface with throughput of 3M bytes/second for connection of
advanced high-apeed devices;
expanaion of the IO inatruction aet; and
facilities for inatruction retry and recovery after failurea.
Technical characteristics for the YeS-4001 channel are:
block-multiplexer channel: .
number 3
throughput, one-byte interface 1.SM bytes/second
throughput, two-byte interface 3M bytesJsecond
byte-multiplexer channel: 7,
aelector subchannels 4~ '
block-multiplexer subchannels 192
selector subchannel throughput 0.5M bytea/second
block-multiplexer subchannel
throughput O.1M bytea/aecond
The main storage unit for the YeS-1060 computer ia built with integrated circuits
with a capacity of 16K bita per package, giving 8M bytes in the standard rack,
which also holds the power supply for main storage. The use of dynamic integrat~~
circuits required the regeneration mode in the CPU storage control unit.
Unit cycle time is 680 na and acceas time ia 520 ns. ~
In the first atage of development of the YeS-1060 computer, basic efforts were
directed toward developing�a processor as a unit reflecting to the higheat extent
the the technical characteristics of the computer.
The YeS-4012 and YeS-4035 units, that were part of the YeS-1O50~computer, were used
as the IO channels. The YeS-3206 main storage unit was built with ferrite cores
with the application of technology developed in the main storage unit for the
YeS-1050. The power supply system also made. use of the secondary power supply .
units for the YeS-1050 c~mputer. Zhis permitted~using series manufactured devices
and units and thereby r~ducing the labor-intensiveness of development and the
period for putting the computer into series production.
The standard conf iguration of the central unita in the computer in the first stage
wa~s: the YeS-2060 CPU, the YeS-4012 multiplexer channel, two YeS-4035 selector
channels, and two YeS-3206 main atorage units with a capacity of 2M bytes..
The standard configuration took up nine racks, including ~our for the CPU and chan-
nel power aupply system.
In the second stage of development of the YeS-1060, the problem of developing a
computer that fully implements tne principles of operation of the second phase of
the Unified Syatem of Computera was solved; this required development of the new
universal channel, the YeS-4001. The problems of developing main storage with in-
tegrated circuits and improving the power supply system by ueing mare powerful
aecondary power suppliea were also solved. The main effo~'ts in this stage were
aimed at de~cloping the YeS-4001 universal channel and develo.ping operating system
versions for the virtual organization of storage.
~ 20 . ~
~ FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2447/02/09: CIA-RDP82-00850R000500444433-6
. FOR OFFICIAL USE ONLY -
In the second stage, i~1 .the standard configuration, the number of IO channels was
also increased to four and main storage capacity to 8M bytes, and the operating,
technical and economic characteristics of the model were substantially improved.
The�atandard conf iguration of central units in the YeS-lObO computer in the seccnd
9tage includes the YeS-2060 CPU, the YeS-4001 unit with one byte-mLltiplexer and
three block-multiplexer channels, and the 8M-byte main storage unit.
The total racks for the central units is four.
During the period of improving the first-stage computer for series production, CPU
facil.ities were developed that support the principles~of operation of the second
phase of the Unified System of Computers, the OS. 6.1 version of the operating sys-
, tem and facilities for recovery af ter failures in the CPU.
The experience of the development, adjustment and operation of the YeS-1060 compu-
ter confirmed the correctness of the chosen direction of staged development of the
model, which permitted development of a computer with high operating, technical and
economic characteristica. Further improvement of the YeS-1060 may involve
reducing the CPU operation step, raiaing the reliability and expanding the
~#.a~gnostic facilities for maintenance. ~
C@EYRIGHT: Izdatel'stvo "Finansy i statistika", 1981 ~
8545
CSO: 8144/0138
- ~ 21
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500044433-6
r'UR UFNt~IAL USE ONLY
UNIFIED SECQNDARY POWER SUPPLIES FOR SERIES-2 UNIFIED SYSTEM COI~UTERS
Irbscow VYCHISLITEL~NAYA TEIQiNIKA SOTSIALISTICHESKIKH STRAN in Aussian No 9, 1981
(signed to presa 21 May 81) pp 19-28
[Article by B, N. Fayzulaye~v, doctor of engineering science (USSR), Ye. I.
Batyukov, engineer (USSR), and Zh. A. [~IIcrtchyan, engineer (USSR)]
[Text] Distinctive features of the second phase of the Unffied System of Computerq
are the wide range of throughput (from tens of thousands to several millions~of
operations per second), improvement of the throughput-cost ratio, use of fast inte-
grated circuits with an enhanced degree of integration, unified principles for the
design of power supplies and the uae of unified secondary power supplies (W IP):
The power supply system (SEP) for the Unified Syatem Computer hardware is part of
the electronic equfpment and largely determines the technical and economic indica-
_ tors for computers in the Unif ied System and their individual units; therefore,
improving the parameters of the power supply system was a major task in developing
the Unif ied System of Computers.
~he power supply system for the Unified System of Computers ia the aggregate of the
- transformers for the industrial power suppTy line, diatributing, regulating and~
_ switching devices, protection and control units, and the secondary power supplies
that form the main part of the power equipment for the~hardware.
The basic requirements imposed on hardware power supply systems for the Unified
System of Computers are: .
modular principle of design;
structural and electromagnetic compatibility between the power supplies.and elec-
tronic and logic units, allowing th~em to be used together in the basic desiga;
- unification and standardization of the.nomenclatuze and design o~ secondary power
supplies (VIP) [SP3] and power control units (BUP); ~
unified system interface for power supplies;
high stability of voltages output from SPS; � .
enhanced immunity to the effect of main power line and cross noise;
provision of the necessary protection and contrnl functions with stable threaholds
- for ~operation; ~
high efficiency factor; and . .
high degree of unification of circLit engineering and deaign solutions with regard
to reducing labor-lntensiveness of development of various aystema anri organization
- of their large series production. � ' ~
~ 22
~ FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500044433-6
FOR OFFICIAL USE ONLY
~'hese requirements were worked out at the stage of developing the first phase a�
- the Unif ied System of Computers and formulated in a series of standards and
methodological instr.uctions.
During development of the units and systems for power supply to the "Series-1"
hardware, much attention was paid to solving the problema of electromagnetic com-
_ patibility between the SPS and electronic equipment and to improving SPS unit
characteristics. Let us discuas these problema in more detail..
Computer power supplies for the f irat and second generations were largely based on
line regulators that are characterized by a low level of pulse and high-frequency
noise; their unit indicators fully meet the requirements of the electronic equip-
ment built with diacrete elements. With the shift from discrete components to in-
tegrated elements in third-generation computers, the~absolute noise immunity of the
element base deciined due to the decrease in amplitudes of the working aignals,
while the SPS noise level increased considerablg due to the use of pulse regulation
methods, the use of key SPS and increased power consumption per atructural unit.
Due to the small working signals of modern integrated circuits [IC] and the small
absolute value of noisE immunity, it became neceaeary to protect computer hardware
from noise in the industrial power line and cross noise generated by the computer
harc~ware itself. The shape and amplitude of external noise dependa on the nature
o�~ ~~e occurrence. However, most dangerous is industrial noise that occurs on the
main power line. Recording and studying it is difficult because it is irregular
and has a wide range in duration of effect and amplitude. Reliable operation of
hardware, in particular the time between breakdown and failure, depends~on the
_ amount and spectrum of noise affecting the power supply and the capability of the
latter to impede the passage of noise to the load to a considerable extent. Frotn
the viewpoint of immunity to industrial and cross noise, the following requirements
on the Unified Computer System po~,aer supply system were formulated as a result of
_ res~arch:
' immunity to the effect of ~voltage deviations in the industrial supply by 80V for
20-30 ms; and
ic~snunity to the effect of voltage spikes with an amplitude of 1000-1500V and
duration of 2-10 microseconds.
= Electromagnetic compati,bility under the conditions of main supply, croes and inter-
nal noise was provided for the SPS as a result of adopting a number of complex
measurea: '
exacting design of the electranic portion of the power supply system and system for
electronic assembly of hardware elements and assemblies; .
use of main line noise suppression filters in the SPS and devices;
use of converters for the primary supply line (motor-generators, staCic converters,
etc.); and
development of a protection system for units and devices.
In first-phase Unified System computer devices, the power aupply system makes up a
significant portion of all electronic equipment (50 to 70 percent of computer bulk).
This is because the size of electronic equipment was considerably reduced compared
to that in second-generation computers owing to the application of IC's, while SPS
consisting of various types of discrete electronic and radio items do not lend
23
~ FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
H'UK UN'N'1(.'IAL USE UNLY
themselves to effective microminiaturization. Because of this, for third-genera-
tion computers, reducing the structural bulk of power supplies and in the first
place increasing the unit indicators of SPS became a basic problem. The increased
requirements on SPS parameters and device packaging deneity, and the need for
raising the manuf actureability and reducing the labor intensiveness in manufactur-
_ ing power supplies required substantial improvement in power supply characteristics
at the stage of developing the second phase of the Unified Systen~.
- At the same time, the structure of the power supply system adopted for the Unified
System had notable advantages: standard interface and modular principle of deaign
based on the developed unified series of unified power supplies (UBP) permitted
development and putting into production power supply systems for all hardware
within a short term. .
The design and introduction into large series pr.oduction of unified power supplies
indicated that that was the only way to enable development of the required quantity
of SPS, install them in the varioua Unified System hardware and achieve high
quality with minimal outlays;
The modular structure of the functionally complete unita with an autonomous syst~~+
for protection and control is a characteristic feature of the design for power
supply systems for the Unified System of Computers. Use of it made it possible to:
organize large series production of the units used in all Unified System hardware;
enable centralized development and high quality fo the units;
employ joint packaging of power supplies and electronic units in the unified base
design for various types of hardware;
reduce the labor intensiveness for development of Unified System.hardware; and
continually improve and modernize power suppliea irrespective of the development
and production of Unified System hardware.
Considering the advantages of the power eupply syatem used in the first phase of
the ihiified System of Computers, listed above, this structure was kept when the
power supply systems and units were developed for "Series-2" hardware.
The main task in developing power supplies for aecond-phase Unified System hardware
was to reduce the size of the power supply system and increase specif ic output
- power of SPS two-three fold compared to that for the first-phase unif ied power
supplies, noise immunity and efficiency factor. The latter is necessary primarily
to ensure normal heat conditions under the conditions of increasing specific out-
put power while keeping the previous structures and to save energy while power
_ consumption is increasing continually in the national economy.
A unified series of unified secondary power supplies was developed and introduced
into series production for second-phaee Unif ied System hardware. These units meet
the requirements formulated above and are built on the basis of high-voltage key
regulators on Che high-voltage side with application of hi:gh-frequency power
s~miconductor instruments manufactured domestically.
The unified nomenclature of the unified secondary power supplies and the modular
compatibility of them with the basic designs of all Unified System hardware have
enabled their wide application in devices used for various purposes.

~ 24
~ FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
FOR OFFICIAL USE ONLY
u
. ~
_ ~
'C w
~ v ~
a~ . d ~
~ ~ ~
~ ~ Q
~ .
~
S.~ .
ab
a~ y^ b ' e0 e0
� s ~ ~
d 3 t~ a� ~ 0
w N � ~ a � - a~ w
L 4 '
~ a ~ ~ ~ ~ ^ ~ ~
vi d ~ a^~ N IJ 1'~.
i~ 1~ JJ o � _
- ~ p, ~ Z � ~ ~ ~ ~
C~ U iYi 4~ o I . " "O
~ C
i~+ ~ Gr+ B ~m a - ~ " ~ ;
oD al
'Lf Q < ,
w
. ~ ~ C 4 p ~
rl ~ ~ ~ ~ . _ Q ~
~
' ~ d < `n" - - - V
~ ~ 3
!.i 3+ ~ " o M ~ . 'C L~
.'7~t QG 4 ~ v ' ~ ~ ~ ~ ~
�C`~, CC ~ n
cQ9', ~ c~ " o ~ - _
~ ~ ~
~ ~ ~ ~ ~ A, a
.c ~ ' a
~ I ~ ~ ~
.c a~
o a~i ~ =
u ~ ~
w~ I ~
~ ~ ~
~u ~ ~ ~ v
L~i CWr� ~ ~ ~ 1 I ~ ' M,~ }~,i
~.1 ~
u f+ ~ ~ N II n
~1~ .
N w r~-+' N
� a, �
~aCi a~ a V ~tiw
w d ~
~ +r i ~ '
.~va~ . ~
m ~ 1 ~
o No
~ ~ I~
~ ad
,.a . .
,a �a.
~
V 'C ~ ~ .
O 1 t7 'C~ O F+ W ~ b c~A d,+C
a a~i r-1 ~ N+~+ at ~ a a+ 3~ a, N u
d ao ..a d u ~tl a ~ ~ i ~ ~ .C
~ C w 3~ b ~~.a ~ 3 w a~ G` u H 3~+ C
ooa~ o ~�+w a~waooa~ omo >+o ~o.o
. ~ 4! ~ O r+ d O 0~ N 00 4J f.i ~ u~ f�+ ,a
~ i.l N~ t~ i-I .~1 �fa J G) O ~ 4) aD ~ 1~ t GD 0! $ 1J
c+ a~ .a ~ ~d o w ~v o w ~ . a ..a a ~e m ~ ~o
d d..+ v ~0 m s+ N,.i d v m t~ eo_ N w ~ ao N~. r+
~o ovw mwa~d d.c.c~+.�+
t-~ ~ w~ > N~+ o~+ cg a3r+ ~ w 3~ y~ C a~
~ ~5 .
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
FOR OFFICIAL USE ONLY
. � o , ~ _ o .
~ o 0 0 0
o o ~ o -
h o ~ ~ o
o o o ~
~ o o p
� - ~ o
0 0 ~ '
o o
0 0 ~ . o
o - ~ r,
- ' c
- o
o 'o
l.~.tiJ
~
~
m. ~ .
,
1 ~ ~
~
~
~
H ~ ~ ~ ~
~
~ .
w ~
~
o ~ .
w~
s+ N C
u u a, 3 r+ e v~ �4 o v~ ~
W ~ O~ d O C
a a?~~~''� yow w a'~a~a+ m~ o N
c~~ d~u r+~+a~ [-~oa~c m~ aU ~a`~i~ ~w~a
~ .o ~ ~ ~ ~ ~o u a w a ~o ~
~ y d ~ ~ ~ ~ ~ W y ~C O ~ ~ ~ O
00 u~ u~~ ~~o~o
V ~ ~~p G O~~ ~ a! ~
> ao a@ V E-+ 3~+ ~ S~~ 3.~i 3 a u 3 w~
. ~ 26 ~ .
FOR OFFIC[AL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
FOR OFFICIAL USE ONLY
Table 2. List of un':Fied secondary power supplies (WIP)
Ca6 x~a (4) no Ilrsa� y~~~
Tw` Y~BH(t ~ A0"' Mace~, aaeru wuta~ ~
ID~p � tpnmwp Kr Kou~� c~n,
uarear eorrr, vrc~w
. ' EC 38M Hr Br/rur~ Hr/aC
2615A EGT'200/B007 1/8 5 90 1 9~0 6~0
2R 45A EC-T200/B005 I J4 8 200 3 12,6 11 ~2
2fs 100A EC�T200/B016 1/2 16 600 3 13~5 11.1
2~4B 8A EC-T200/BOl9 1J16 ~ 1,5 50 I 11~2 12~8
~ 5B 8A EC-T200/BOl8 1/16 ' 2,5 80 ! 23,5 16~0
2~4B 80A EC�'~'200/BO15 1/2 l8 600 3 13,0 10~T
SB 1SA EC�T200/B022 . 1/8 S 9b 1 22,T 9,0 �
58 40A EC-T200/B004 I/4 8 3$0 3 28,1 25~0
~ SB 90A EC�T200/B002 I/2 11,5 ?00 3 3p~4 39.0
, 5,28 15A EC�T200/B006 1/S 5 140 1 23,6 15~6
5~2R 40A EC�T200/8003 I/4 8 3$0 3 29~3 26,0
5~28 80A EC�T200/B001 1/2 11,5 T00 3 25~4 36~0
6,3B SA EC-T200/BO17 1/16 I,5 6U 1 18~5 l7~5
. 12B 10A1 EC-T200/H002 I/4 8 170 1 16~9 IS
12,66 2A EC�T200/BOl2 1/16 1,8 60 1 14,8 14~0
12,66 ZOA EC�T200/B013 ' 1/2 !0 450 3 (7,0 25,2
!SB 6A , EC-T2Q0/H001 I/4 8 .170 l 12,6 11,0
20B ZA EC-T200/8011 I/ 16 1,S 6~ 1~ 23,b ~7l ~2
206 4A EGT200/B024 I/8 5 125 I 24,2 16,0
- 20B ISA EGT200/B009 1/2 20 600 3 20,2 15,0
~ 276 2A EC-T200/BO10 1/16 1~5 80 l 31,7 30,0
2TR 2A EC-T200/BO10 1/!6 1,8 80 1 31,7 30~0
27B 3A EC-T200/B025 1/8 5 l25 1 24,5 16~0
27B 13A EC-T200/B008 lJ2 20 600 3 23~7 17~5
'~B 8~ EC-T200/H004 1/2 11,5 50 1 26~0 33,3
56I10A EC�1052/H002 1/2 - - 3 40,0 40~0
f a6apxrs~e pa3e~ep~ Ao.lefk rxnoeo~ naHenx EC ~3BM, e~x: �
(10 ) 1~2-386X 178X215; .
1/4 -1T9X 178X215;
1/8 - 89 X 178 X 215; ~
. 1/16-44X1T8X215. '
Key:
1. power supply type, B= Volts, A=~inps,
2. model number, EC s YeS, T= T, B= V, H= N .
3. dimensional share of standard Unified System panel
4. weight, kg _
5. power consumption, watts
6. feed line, number of phases
7. specif ic power ~
8. W/dm3 :
9. W/kgf _
10. dimensions of share of standard Unifiec~ System panel, m~n
27
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500040033-6
r~K urr~~~A1. u,~ u1vLY
There is a trend to uaing SPS with traneformerlese input in both domestic and
foreign development of computer hardware power supplies. Table 1 shows the applic-
ability of the different principles for electrical conversion and regulation in SPS
for computer hardware and other radioelectronic apparatus. Analysis of acientific,
technical and patent literature indicates that the majority of foreign f irma tend
_ to use the method of pulse-width voltage regulation; in the process, used as output
stages are transistor bridge and half-bridge variable inverters with transformerless
, input. In a number of cases, single-cycle converters and line voltage regulators
have been found to be more efficient for SPS with output power of less than 100W.
Table 2 lists the unified secondary power supplies developed and put into large
series production. Characteriatics of the unified power supplies and unified
secondary power supplies [SPS] are:
UP 3 SP S :
Total instability, % 5 3+4 ~
Retention of output voltage, microsec. 0.1+1 10+30 ~
_ Efficiency factor, % 35+50 65
Specific output power, W/dm3 10+12 30+40
R dissipation 1.0+1.85 0.55 ~
R load -
From this information, it can be seen that with the design of the SPS, specific
output power was increased two to threefold, amount of total instability of output
voltage was reduced 1.5-fold and the efficiency factor was raised 1.3-2 fold which
permitted keeping the total heat release within the supply casing.
The basic indicatora of the key SPS with transformerless input, uaed in the central
part of the computers, in epecific indicators are similar to the SPS of the
socialist countries.
It is interesting to compare the basic apecific indicators for the power supply
systems and units for the first and second phases of Unified System hardware. The
first-phase sqstem was built on the base of unified power supplies (UBP) with trans-
former input, developed by using linear and combined principles for converting the
electricity. The most powerful unit in thie seriea, the UBP SV/60 A, was designed
according to the structural scheme with a variable thyristor inverter with trans-
formerless input and 1 kHz conversion frequency. Introduction in Unified System
power supply systems of UBP series units afforded their electrical and structural
compatibility with Unified System hardware; in the process, the ratio of power sup-
ply size to total computer device aize was 0.5 to 0.7.
Using modern methods of high-frequency conversion and pulse regulation of electri-
city in power aupplies for the second phase of the Unified System of Computers per-
mitted reducing the size taken up by the power supply to 0.3-4.5 of the total size
of the electronic equipment. Thus, for example, the power supply for the YeS-1060
computer processor with UBP type units (thyristor inverter with 1 kHz conversfon
frequency) took up 2.5 atandard racks; switching to the SPS permitted placing the
corresponding power supply system in just one rack. Thus, the change from UBP to
the more economical and smaller SPS with tranaformerless input and conversion of
electricity at the increased switching frequency of 20 kHz permitted reducing the
28
FOR OFFiCIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED F~R RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
FOR OFFICIAL USE ONLY
corresponding size of the power supply system two to threefold. This made it poe-
sible to place the electronic portion of the hardware and power supply unita in one
cabinet and to meet the requirement on electromagnetic and structural comFatibility
between the power supplies and Unified System computer hardware. The increased
requirements for electromagnetic compatibility were met, including through intro-
duction of motor-generators in the computera. '
Structurally, W IP [unif ied aecondary power supplies]~ are units (see photo) with a
width equal to a main panel. The height and depth of all units are identical. ,The
maximum dimensions of theae SPS correspond to the dimensions of a half panel, and
the minimum to 1/16 of a panel. Placed on the front panel are controls for the
continuously adjustable level of output voltage, for switching to the mode of pre-
ventive change in the levels of output voltage by �5% of the rated value, and a
warning light. The unit rear panel has a connection for the control and protection
circuita, a connection for input of the main supply voltage, and an output for the
faedback and output voltage circuits for low-power aources. Sources with dimen-
siona of 1/4 and 1/2 of a panel have special terminals to feed the output voltage.
~
a
.
General view of the unified secondary power supply
SPS built by using high-frequency conversion and pulse regulation on the high vol-
tage side (300-600V) with pulse regulators and voltage conv~rters are promising for
further development of the Unified System of Computers. In designing future power
supply systems and units, consideration should be given to a 10-15 fold increase in
the next five to seven years of the denaity of packaging of logic elements in third
and fourth generation Unified System hardware built with matrix LSI circuita and
LSI storage. This trend has to be conaidered in the design of SPS and power supply
systems to ensure their electrical and technological design cocapatibility with the
- 29
� ~ FOR OFFIC[AL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
. r~K urr~~iwL u~~ uNLY
electronics of the processing devicea. In the process, advance development of the
element base is becoming a topical question. The design of this base must be
developed under a cortanon plan coordinated emong the countries that participate in
. the agreement on develo~nent of the Unified System of Computers.
Achieving specific indicators of 120-150 3 required for the near future and in-
~ dm
creasing the efficiency factor to 0.15-0.8 primarily involve developing a highly
efficient power element baee:
transistors with working voltage to 1-1.5kV, switching currents to 10A, threshold
frequencies of 10-20 I~iz and turn-off time of 0.25-0.5 microsecond;
variable thyriators with turn-off time of 1-3 microaeconds; ~
rectif ier Schottky diodes with working currenta of 30-100A;
electrolytic capacitors with high indicatore for specific electrical energy
2
(5-50 F~3 for compacitors wi~h woFking voltage to 16V); and .
dm
integrated circuits for SPS control and protection circuits.
Using methods of high-frequency converaion and pulse regulation of electricity and
development of a new element base will allow creation of highly efficient power
supply systems for future computers in the Unified System based on SPS with puls
regulators and high-frequency convertera of main eupply voltage with
transformerless input. .
COPYRIGHT: Izdatel'stvo "Finansy i statistika", 1981
8545 ~
CSO: 8144/0138
,
. 30
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2407102/09: CIA-RDP82-00850R000500440033-6
~ FOR OFFICIAL USE ONLY ~
~ DATA PREPARATION UNIT DEVELOPMENT IN BULGARIA .
- Moscow VYCHISLITEL'NAYA TEKHNIKA SOTSIALISTICHESKIKH STRAN in Russian No 9, 1981
(signed to press 21 May 81) pp 28-32 ~
_ [Article by T. A. Topalov, candidate of engineering acience, (People's Republic of
Bulgaria)J
[~Rxt] Bulgarian epecialization in external magnetic media storage units allowe
Bulgarian specialists to salve a nwaber of problem~ involving the design and pro-
~y~~ton of data preparation uni::s (UPD) LDPU] using magnetic media. In doing so,
oqqplex engineering problems 1,lave to be solved: developnent of electric motors
with special characterietics, magnetic heads, contactless switching elements,
methode of magnetic recordi~~g and others. Also, development of DPU~a requirea the
optimal solution to a numbo~r of problems characteri~ing the "operator-device--
medium-computer" syatem, f~r example, chooaing the medium for data preparation aad
entrq to a computer system.
The DPU's developed and manufactured ir~ Bulgaria use standard magnetic tape with a
width of 12.7 mm, recording density of 32 bits/nmt and a floppy disk with a diameter
of 203 man. The key-to-tape units developed earlier than others have�adequate ca-.
pacity and f ine quality and are produced in the socialist countries. They are com-
patible with computer external storage unita. The floppy disk ia compact and re-
liable. Key-to-diak unita are characterized by random accesa and uaing them ie
promising for mini and microcomputer systema. � ~
Uaed in developing DPU's is the experience of optimizing the data preparation pro-
cesa gained over many years of operation of key punch units, since the new DPU~s
must be suff iciently close in a syatems aense to the key punch units. Otherwise,
the new uaits would not be applied in the data proceasing process already estab-
~ lished.
The first key-to-tape un~.t developed in Bulgaria was the YeS-9002. Compared to key
punch units, work efficiency is raised conaiderably by the high apeed of aoftware
and manual control of the data format input or v~rified from the keyboard, charac-
ter display of current information, quiet operation of the unit and simple selection
of the mode or function. ~1e YeS-9002 gives the user the capability of achieving
high economic indicatora beceuse of its high reliability, longer amortization time,
time-saving computer entry operations, verification of data file preparation (which
requirea apeciaY devices with paper media), and the capability of reuaing tapes.
Produced over several years, the YeS-9002 i9 in succeseful uae in a number of
FOR UFFIC[AL USE .ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500040033-6
FOR OFFICIAL USE ONLY
countries. Modifications have been made to thia unit, for example, the
YeS-9002.02 can print out information from amgnetic tape.
The YeS-9004 ie an updated version fo the YeS-9002; it has substantially improved
functional, design and operating characteristics. It has a current data block dig~
play, based on a CRT with television scan or the "Elektronika VL100" and "Yunost'
603" consumer television receivers. This unit is very flexible with respect to the
character set and recording density; for example, it is easy to modify the YeS-9004
for a recording density of 8 or 22 bits/mm. The uaer has the option of having the
unit equipped with the functions for modulo 10 sum, modulo 11 swn and total sum of
all fields..
The use in the YeS-9004 of circuits with an increased scale of integration, connec-
tors with improved characteristics and a more successful design for the tape drive
permits a substantial increase in the indicator for mean time between failures.
The unit has a detached keyboard and smaller dimensions and weight. Zhe capability
of connecting a number of IO devices extends significantly its sphere of applica-
tion.
The YeS-9004 has these modifications:
the YeS-9004.02, with output to the YeS-7187 printer;
the YeS-9004.03, with input from the IZOT 6001 card reader; ~
the YeS-9004.04, with a modem for data exchange over telephone lines;
the YeS-9004.05, with input from a perforated tape reader;
the YeS-9004.06, with output to perforated tape; and
the YeS-9004.07, with floppy disk storage.
The following table compares the YeS-9002, YeS-9004 and the irIDS Data Record 6441
used in capitaliat countries.
The devices operate in these modes: data entry, verification and retrieval; pro-
gram entry and verification; and buffer storage input, verification and readout.
Parameter I~S 6401 YeS-9002 YeS-9004
buffer storage ferrite integrated integrated
element base discrete IC~a IC's
display bit character block
tape drive presaure 1 roller 1 roller
roller
recording method NRZ1 . NRZ1 NRZ1
density, bit/mm 32 32 8, 22, 32
temperature range, �C 10-38 5-40 5-40
weight, kg 122 64 48
dimensions, mm 794x743x1092 582x641x583 481x396x483
data validity,.bits 10~ ~ 108
MTBF, hours 500 1000
32
FOR OFFICIAL USE ONLY .
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
FOR OFFiCIAL USE ONLY
In recent years, a complex of DPU's using floppy disks has been developed and is
being aseimilated in Bulgaria. Testa on the SM-6910 key-to-floppy disk unit have
been successfully concluded. Prototype series of the YeS-9112.01 key-to-f loppy
disk unit and the YeS-9113 f Loppy-to-tape converter is being completed. All these
units are built with microprocessor circuits as the element base. They are modular
and have a common external shape.
The SM-6901 includes a rack with keyboard, electronic unit and power supp.Ix, cabinet
with two floppy disk storage units and power supply, and a video monitor.
SM-6901 characteriatics:
element base LSI circuits ~
ISO compatible yes
programs 10
data validity 10~
MTBF, hours ~ 500
weight, kg 133
The SM-6901 performs the following functions: entry, processfng, checking, retriev-
al by address, retrieval by content and end of data; initialization; count of total
aa~ check sum; copying of content of floppy, file, specific block; operation with
cl~a~n of programa; statistics of operator work.
The YeS-9112.01 differs from the SM-6901 only by the codes of the characters in the
DKOI [decimal interchange code]. Data prepared on the YeS-9112.01 and SM-6901
units can be entered into a computer directly from a YeS-5057 unit ;.or Unified Sys-
tem computers and from external f loppy disk storage units for the System of Small
Computera.
The YeS-9113 converter can perform all the functions provided~by the SM-6901 and
YeS-9112.01 units and a number of additional ones:
finding a tape mark;
finding an end of file on tape;
retrieval by content on tape;
movement of tape forward;
movement of tape backward;
copy from floppy to tape;
copy from tape to floppy;
copy from floppy to tape on a specific condition;
copy from tape to floppy on a specific condition;
key-to-tape recording; and
deletion of data blocks on tape. '
The YeS-9113 includes a rack with keyboard, electronic unit and power supply,
cabinet with two YeS-5074 NGI~ [floppy disk storage units], cabinet with SM-5300
NML [magnetic tape storage unit] and a video monitor.
1~e use of microprogram control in these units and external storage units with ran-
dom access leads us to expect a further increase in operator productivity compared
to that obtained with the YeS-9002 and YeS-9004 units.
33
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
rvK uh~r'tc:iAL U5E ONLY
_ The YeS-9003 multiconsole key-to-tape system handles the complex problems involving
data preparation and entry of files into a computers. Use of a general-purpose
microprocessor in the system determines the implementation of a set of complex func-
tions. The YeS-9003 system is built with the following hardware:
IZOT 310 CPU (32K 12-bit words storage);
SM-5400/O1 NhID [magnetic diak storage unit] (6M byte capacity, 1500 rev/min);
SM-5300 NML [Magentic tape storage unit] (recording speed is 32 cm/s, density fs
32 bits/mm, method is NRZ1);
IZOT 0232 operator console;
YeS-7186 mosaic serial printer (rate: 180 cps, 132 char/line);
YeS-0101 keyboard ~16 keys); and
video monitor with display of 128 characters (4 lines of 32 characters each).
However, the high cost of the system units make it noncompetitive with standalone
buffer storage units in configurations using less 8 to 10 consoles.
Studies have shown that operator productivity is increased from 40 to 120 percent
when these units are us:d, compared to key punch operations, and data preparation
expenses, which account for 30 to 50 percent of total computer operating cost, are
substantially reduced.
Irrespective of the fact that on-line proceasing ie more pramising, data preparation
in the batch mode will predominate in a number of cases. Naturally, for this it is
necessary to continually raiae the technical and economic indicators of the hardware
employed. Efforts are undeiway in Bulgaria primarily in two directions: optimiza-
tion of the operator-device interface by expanding functional, editing and proces-
sing capabilities and by improving operating characteristics, especially hardware
reliability, by searching for new design and engineering solutions.
COPYRIGHT: Izdatel'stvo "Finansy i statistika", 1981
8545
CSO: 8144/0138
~ 31~
FOR OFF'ICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00854R004500040033-6
FOR OFFICIAL USE O~ILY
YES-1045 COI~UTER, CHARACTERISTIC FEATURES
Moacow VYCHISLITEL'NAYA TEKHNIKA SOTSIALISTICHESKIKH STRAN in Russian No 9, 5~381
(signad to press 21 May 81) pp 32-41
[Article by A. T. Kuchukyan, T. Xe. Sarkisyan and I. B. I~Ilcrtumyan, all candidates
of engineering science (USSR)]
[Text] Zhe YeS-I045 is a general-puxpose compu.ter with med~ium throughput designed
iCp atolve a broad range o~ scientific, technical, economic, tnfornnatio.nal and logic
P.~ab;~,.ems �
The logic structure of the YeS-1045 meets the r.equir~ments imposed on the logic
structure of the second phase of the Unif ied System of Computers.
The YeS-1045 computer hardware i's modular which allows a user to design specialized
computer systems correaponding to the purpose o~ the syatem being created to the
highest extent without making.any changes to machine structure and design.
YeS-1045 computer hardware can be classif ied as. standard, part af any system con-
figuration, and optional, delivered upon custome~ requeat.
YeS-1045 standard facilities are:
facilitiea for proceasing in the basic and exeended control modes; ~
the~Unified Syetem universal instruction set;
dynamic address translation [DAT];
ir?direct data addressing in channel;
instruction exeaution retry; ~
� monitor facilitiea; ~ .
program event recording [PER]~;
fast 8K-byte buffer storage;
byte-multiplexer channel 0; and ,
' block-multiplexer channels 1 and 2.
YeS-1045 optional facilities are:
floating-point instructione;
direct control facilities;
facilities for organizing a dual processor system;
block-multiplexer channe,ls 3, 4 and 5;
a second byte-multiplexer channel (replaces block-multiplexer channel 4);
FOR OFF[CIAL USE ~N~Y
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500040033-6
� ruK urr~tc:iwL u~~ UNLY
one ~r two channel-to-channel adapters;
logic repeater; and
configuration panel.
The basic set of the YeS-1045 computer is placed in three standard Unif ied System
racks (processor and IO channels, main storage and pcwer supply). The channel-to-
channel adaptera and~logic repeater are placed in tfie same racks.
In the YeS-1045 computer, CPU control is hardware-microprogram with a machine step
of 120 ns. YeS-1045 throughput using a Gibson mix for scientific and technical
problems reaches 880,000 operationsJsecond; for economic problems using a GPO mix
it is 540,000 operations/second. ~ .
The CPU includes a microprogram control unit (BMU), instruction fetch and interrup-
tion servicing unit (BVK) and operations unit (BO).
The microprogram control unit is designed to control CPU and IO channel operation
by means of microprograms stored in control atorage (UP). Machine control storage
is made with bipolar tra~.sistors and consists of two parts: permanent control
storage (UPP) and loadabl~ control storage (UPZ).
Permanent control storage (addresses with the ranges of 0-6 and 7-8K words) is
designed for storaga of CPU.and IO channel control microprogratns; loadable control
storage (addresses 6-7K worda) holds microprogarms for console operations and. diag-
nostic mic~oprograms. Loadable control.storage can also be used to~introduce
special operations in~solving specific user problems on the YeS-1045. Micropro-
grams are loaded from the ML-45 magentic tape cassette storage console unit.
The instruction fetch and interruption servicing unit is designed to prefetch,
store and partially decode instructions, fetch operands and service interruptions.
The version adopted in the YeS-1045. computer processor is an instruction fetch unit
with buffering which enables overlap of instruction fetching from the inetruction
buffer storage (HPK) ~or main storage with execution of instructions in the CPU.
_ Instruction buffer storage is made with integrated circuits and has a capacity of
32 8-byte words. It is designed to hold 64 words of instructiona, prefetched from
main storage, with the capability of their rapid fetching and iu;;ving into the CPU
to fill three instruction buffer registers. These registers are filled af ter the
contents of two of them are used. Instruction fetch time with this scheme is
actually determined by the transfer from one buff er register to another. Because
of this and the availability of the instruction buffer storage,. instruction fetch
- eff iciency is appreciably enhanced since practically the majority of problema make
use of blocks of instructions with cyclic repetftion within these blocks. ~
The operations unit consists.of four- and one-byte data paths; it is designed to
execute a broad set of arithmetic and logic operations used in the Unif ied System.
A special high-speed processor-accelerator is provided in 'the operations unit to
speed up execution of certain "long" arithmetic and logic operations. Executed in
the accelerator are all multiply operations (except extended), binary-decimal trans-
lations, packing and unpacking, all primary shif t operations, and some move and
store operations, 25 operations in all. Th~nka to incorporation of th~ accelera~tor
in *_he operations unit structure, CP�U throughp~.it increased an average of 15-17 per-
cent while equipment bulk increased only 5-6 percent. .
36
FC~R OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2407102/09: CIA-RDP82-00850R000500440033-6
FOR OFFICIAL USE ONLY ~
- The main storage control unit (BUOP) organizea data exchange between the CPU, IO
' channels, array processor and main storage; provides for rapid access of frequently
used data and instructions from the appropriate blocks of buffer storage without
- re=erencing main storage; converts virtual addresses into real; determines and
signals violations of storage protection; and organizes a common extent of maitt
storage of two computers in a dual processor system.
Buffer storage (BP) with a capacity of 8192 bytes is made with integrated circuits
with data access time equal to a CPU step. The entire 16-byte block of data is
kept in buffer storage when the CPU references main storage for some part of this
block. Thanka to the algorithm used to fill buffer storage, the probability of
finding the next opexand in the buffer ia about 75 percent, which reduces the
average time for fetching data used most often by the CPU.
Virtual addresses are translated into real by using the special apparatus, the
dynamic address translation facility (DPA) and the corresponding segment and page
- tables. This process is speeded up (intermediate references to main storage are
excluded) by the translation-lookaside buffer (BBP), where the real addresses are
written for the pages last referenced. Fetching a real address from the translation-
lookaside buffer [TLB] d~es not lengthen the main storage reference operation.
Provided in the YeS-1045 computer is the capability of simulataneous storage in the
TLB of the results of translation tor three blocks of virtual storage, which allows
avoiding a complete purge of the TLB when a program is r~placed. The DAT hardware
- facilities provide the capability of using up to 16M bytes of virtual storage.
The main storage adapter provides the interface between the main storage control
r ~unit and main storage, directly controls main storage operation in the read or
write modes, and detects errors when data is read from main storage by using the
modified Hammin; code checking apparatus. Single errors are corrected by usir.g the
correction circuits; corrected data is sent from the adapter to the data exchange
control unit. In the process, the data in main storage is not corrected.
Storage protection facilities provide protection of individual zones of main stor-
age from unauthoriaed access. The protection mech3nism provides protection in the
- store mode alone, and during both the store and read modes. A 2048-byte block is
the protected unit in.main storage. The main hardware protection facility used is
the high-speed storage with a capacity to 4096 bytes, which provides protection to
a maximum of 8M bytes of main storage.
There are two types of channels in the YeS-1045 computer: byte-multiplexer (BTMIC)
that can operate in the byte-multiplexer and burst modes; and block-multiplexer
(BLMK) that can operate in the block-multiplexer and selector modes. The IO chan-
nels in the YeS-1045 are controlled by the combined hardware-microprogram method.
Data is exchanged through the IO interface under control of channel hardware facil-
ities concurrently with CPU operation. Data exchaneg between IO channels and main
storage and processing of control information are performed by CPU facilities under
microprogram control. IO channels share control storage apparatus wiT.h the CPU.
During operation of IO channels under microprogram control, execution of the cur-
rent instruction in the CPU is halted, and the MP [array processor] facilities and
- other CPU facilities are made available to the IO channel. When several IO channels
request microprogram service as the same time, service is given based on the hard-
ware priority scheme.
37
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2407/42/09: CIA-RDP82-40850R000500440033-6
FOR OFFICIAL USE ONLY
Control words of the active subchannels for all aix IO channels are stored in a
section of local CPU storage with a 64-word capacity; each channel is allocated
8 words. The Iast 16K-byte area of main storage is used to store control words for
passive subchannels. Up to 10 devices at a distance up to 60 m can be connected
physically to each channel. A logic repeater is made part of the machine to extend
- the distance or number of devices connectable through the IO interface. Throughput
of each byte-multiplexer channel is 47K bytes/sec in the multiplexer mode and up to
110K bytes/sec in the burst mode. Without considering data chaining, block-multi-
plexer channel throughput ranges from 0.5M to 1.SM bytes/sec as a function of the
number of them. Total throughput of a~l IO channels is SM bytes/sec.
The YeS-1045 computer has a highly develoepd checking and diagnoatic system. The
main components of this complex of facilities are hardware-microprogram recovery of
179 instructione when random failurea occur, microdiagnostics that enable checking
all computer facilities in the static and dynamic modes, central checking and mea-
surement of power supply, and automatic preventive checking and diagnostics of
faulty integrated circuits in TE2~s [standard exchange cards]. ~
Yes-1045 computer recovery facilities enable continuation of computer operation when
single errors occur in main storage, bypaseing the consequences of failures in the
CPU and channels through hardware-microprograra retry of operations, and recover~� f
the operating system with the aid of software facilities.
The YeS-1045 computer can also operate with a reduced amount of buffer storage
when a part o~ it fails.
- When the check circuits detect a machine error, control synchropulses are blocked,
- i.e. execution ~f instructions at all levels of overlap is terminated, and hardware
logging of computer status is performed in the control console storage (PPU) with a
capacity of 512 bytes. Hardware logging in the control console stcrage is performed
during the first and eighth retry attempts. Control console storage allows pre-
serving the computer.status at the time of the malfunction when servicing of inter-
~ ruptions from the check circuits is suspended, and the aufficiently large capacity
of it provides the capability of having information on the status of practically all
machine flipflops and registers. After hardware logging is completed, CPU and chan-
nel registers are purged, the possibility of retry of the interrupted operation i~
- analyzed. The retry method depends on the type of operation, computer status at the
time fo the error and the type of error.
Checking for change in initial data is provided for in microprograms of instructions.
When the data is changed, the microprogram branches to alternate execution of the
instruction. Preservation of initial data ia provided for to recover computations
during a malfunction. The following facilities are used for this in the YeS-1045:
blocking of change of the first operand in RR and RX format instructiona upon an
error signal, which permits dispensing with storage of initial data;
~ storage of address of second operand in work area of local storage and recovery of
it during retry, which permits dispeneing with storage of base register contents;
storage of the changed part of the first operand with subsequent recovery of it
during failure in decimal artthmetic instructiona;
storage of Lhe count of successfully processed bytes in RR format logic instruc-
tions with continuation of processing upon failure with retry of proceasing of one
byte;
- 38
- ~ FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED F~R RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
FOR OFFICIAL USE ONLY
checkpoints in microprograms of certain instructions, thanks to which upon retry,
the successfully executed part of the instruction is not repeated.
In tiYe YeS-1045 computer, hardware-microprogram recovery of IO instructions is pos-
gible to a certain threshold, determined by the start of communication with a da-
vice through the interf ace. If the threshold of hardware-mi~roprogram retry of an
IO instruction is passed, an IO interruption is executed and the usual and expanded
status word containing the data specifying the Iocation and conditioa of the error
is etore~. After this, recovery is performed at the program level by usfng fihe
CCH and ERP programs. The CCii program at the �ame time generates a record in the
log file.
In the YeS-1045 computer channels, retry of channe~. coartnands is also provided for
- at the request of a device equipped with the special facilities for iesuing tfiis
request to a channel. Thanks to this, when an error occurs in a device during
execution of a chain of commands, the channel supports reis3ue of the last
command upon request from the device.
Hardware diagnostic facilities include the service adapter and the ML-45 console
storage unit. The service adapter enables finding and reading of data from the
co~sole storage unit, accumulation of data, decoding of diagnostic operations,
ex~cution of microdiagnostics and comparison of computer status with a reference
~'~'atidard. To simplify the loading path, data from the console storage unit goes to
tti~ service adapter in series code aad after its translation into parallel goes to
the CPU circuits. This small amount of hardware facilities,for the diagnostic unit
(seven TEZ~s [standard exchange cards] for the service adapter and five for the
electronic part of the ML-45) substantially increases the operational reliability
and repairability of the YeS-1045 computer. To check the hardware part of the com-
puter, the service adapter makes use of data paths and microoperations that exist
in the computer.
Yes-1045 computer diagnostics consist of three stages.
At the first stage of the microdiagnostics, control of the halted CPU fs monopolized
by ttie service adapter. The CPU executes individual microoperations upon orders
from the service adapter. To poll computer status, the service adapter uses data
aerialized with the aid of a multiplexer, i.e. data on the status of computer regis-
ters and flipflops in the form of a serial stream of bits. Each data byte in the
- serial stream has its own addresa and during the microdiagnostic procedure is
fetched from the general stream to the data register in the control console; signals
output from here go to the circuit for comparison with the reference value kept in
the diagnostic register in the service adapter.
The standard procedure in first-stage microdiagrtiostics includes:
loading of a microinstruction from the console storage unit into the data register
in control storage and execution of it. Test sets are defined by the apecifying
field of the microinstruction. Executed in this mode are the microinstructions
needed to check the functions of the assembly being tested;
reading from the PNK [expansion unknown] of the address of the register (in the
general stream of serialized data, the content ~f which is compared with the stan-
dard, and the reference srandard itself;
39 ~
- FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/49: CIA-RDP82-00850R040500040033-6
rux urNU;iAI, u5~: UNLY ~
~
comparison of the result with the reference etandard in the comparison circuit
SERAD [expansion unknown];
halts with display of the halt number on the control console lights. 1he halt num-
ber is uaed to locate the commenta in the microdiagnostic printouts that contain
the identif ication of the improperly functioning flipflop, list of faulty TEZ'g
Lstandard exchange carda] and recommendations for locating the maifunction.
In the error halt :~tatus, service personnel can obtain additional information on
the system indfcation.
Comparison of the low-order bits of the address of control storage with the
ref erence is provided �or to check the microoperationa of analyses in the diagnostic
apparatus. This permita identifying the microoperations of analyses of modiffcation
of control storage address bits that enabZe branching according to the various com-
puter states. First-stage microdiagnostics also check the facilities for manual
control fo the YeS-1045 coneole by simulating actiona of keys and switchee. Con- ~
venisnt facilitiea for test cycling and microinstructions in the microdiagnoatic
test are provided to facilitate finding malfunctions in the diagnostic unit.
Checked in the first stage of microdiagnostics are all hardware facilities of the �
CPU, channels, storage control, some fo the storage units, coneole typewriter ar~
chaanel-to-channel adapter. Apparatus is checked in microoperations at the rate
- specified by the rate of reading of data from the console storag~+. unit, which~ simu-
lates the "rapid" step mode more closely than the actual machine iate of executing ~
microoperations. Therefore, tests in the first-stage microdiagnostics cannot
identify some errors associated with the timing diagram of operation of computer
~ unita. ~
Second-stage microdiagnostics, making use of loadable control storage checked in
the first atage, identify errors associated with the timing diagram. In this stage,
microdiagnostics are performed under control of the computer itself at the actual
machine rate the following way:
a diagnostic microprogram is input to control loadable storage from the conaole
storage unit and receives control af ter loading;
the diagnostic microprogram checks computer assemblies in the diagnoatic mode, i.e.
at the actual machine rate; after storage of the check results, control is returned
to the service adapter for analysis of the check results by comapring them to the
reference standard under control of the service adapter.
The set of microdiagnostic etsts ia fixed and designed for the maximum configuration.
~The type of channel connected under, a specific number or its availability in the
system is determined by microdiagnostics by polling the special facilities of the
channel that store this information. When aome channel ia not used in a system,
the microdiagnostic teate for this channel are automatically skipped. The amount
of main atorage in its microdiagnostics is determined by the condition of the '
invalid address that enables checking the amount of storage that is available in
the system. Increasing the amount of storage with such organization of the micro-
diagnostics.does not require changing it.
The f irst two stages of microdiagnostics permit checking and localizing a malfunc-
tion in the YeS-1045 computer hardware facilities, including all storage devices.
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
FOR OFFICIAL USE ONLY
The third s'tage of computer diagnostics ie perfQrmed by using the facilities of
functional (software) diagnostics. The purpose of tttis atage is to check and detect
improperly functioning system facilities: the system of instructiona for external
devices, external device control units and concurrent operation of devices. Ttie
YeS-1045 computer functional diagnostic facilities are teh basic test, the set o�
test-sections executed under control of the diagnoatic monitor, and the set of'
autonomous diagnsotic testa. Checking the functional facilities of the system
begins with a check of the nucleus of ttie instruction system by using the basic
test. vardware facilitiea in the diagaostic unit permit loading the basic test into
main storage from the console storage unit. The basfc teat checks the group of CPU
instructions, IO instructions and the paths far communication with the standard
external devices-tape drives and the typewriter. Further checking of system func-
tiona~l facilities is performed in the teat-sections of the KPTO [maintenance program
complex].
- Computer operating experience indicates that localization of failures in the power
aupply system and fans, places of excess heaC, and preventive checking of the volt-
age of secondary power supplies, the number of which is aeveral dozens, require
considerable time. To obtain the required paremeters of operating reliability in
the YeS-1045 computer, there is a system for checking and diagnosing power aupplies
~hich performs the following functions:
au~omatic checking and measurement of output voltages from power supplies; .
ai~~oinatic and manual setting of preventive levels of power supply voltages (WIP);
lo~alization of the place and type of emergency malfunctions, such as hot spots,
fan failurea and power control unit (BUP) failure with audible and light warning
signals.
In measuring voltages output from W IP [unified secondary power supplies = USPS],
the measuring switch connects the outlet of the selected USPS to an analog-to-
- digital converter of a digital voltmeter; information from this goes to a comparisan
circuit.
When USPS voltage deviates beyond tolerances, a signal is generated that goes to
the display.
Preventive five-per-cent levels of USPS voltages are set either automatically
through the 12-channel tnterface from the diagnostic unit or maunually from the sys-
te~ console. The USPS address goes to the decoder; after decodir:g it controls the
switches for the preventive check, ~to the input of which also com~:s the condition
of the level of +5%.
An audible alarm signal is generated and the address of the malfunction is displayed
when overheating occurs, fans stop or the power control unit malfunctions.
A basic problem in operating computers with incomplete ZIP [kit of apare parts,
tools and accessories] is repair of faulty TEZ [standard exchange cards = SEC].
Manufacturers have to set aside up to five percent of their capital, investing it
in ZIP's, and the cost of storing them in warehouses is rather high. Also, the
volume of repeated return of faulty SEC's has increased. It is estimated that
ebout 50 percent of the SEC's is returned to the user without repair since no mal-
function is detected in them. It is cheaper to repair the SEC's on site. This is
poasible with the YeS-1045 computer because of ttle availability of the integrated
!~l
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2407102/09: CIA-RDP82-00850R000500440033-6
N'UN UN'FICIAL USE ONLY
autotester, diagnostic tests, referencea that are part of the operating documenta-
ti,on and the PET-1 tester. The YeS-1045 computer autotester uses the facilities of
diagnostics and the conaole storage unit to test SEC~s. The advantagea of the auto-
tester are its low cost (it has only three SEC's), simplicity and convenience of
maintenance, detailed diagnostic documentation, quick recovery, low demands on the
classif ication of maintenance personnel and the capabflity of reducing ZIP~s.
TY ial operation of the YeS-1045 has indicated that the checking and diagnostic
facilities described above substantially raise computer operating efficiency and
may be recommended for application in other machines.
The YeS-1045 computer structure allows setting up dual processor and multimachine
systems. Dual processor systems are established on the basis of a common extent of
main (up to 8M bytes) and external storage for both processors and peripherals
operating under the control of a single operating system. A special configuration
panel combined with the appropriate'equipment provides the capability of controlling
the modes and configuration of the dual procesaor syatem.
Multimachine systems are set up by complexing at the channel level by using the
channel-to-channel adapter, direct control facilitiea, a common extent of external
storage and common peripherals; each computer operates under control of its own
operating system.
The array processor (MP) is an option for the YeS-1045. It is a specialized compu-
ter connected to the YeS-1045 computer ae the seventh channel. It is used to pro-
cess numeric series encountered in problems where repeated operations have to be
performed on a group of data (series, arrays), thereby relieving teh load on the
CPU.
The array processor is organized for parallel proceseing of data in the CPU and
_ array processor. It has three units: control, arithmetic and microprogram control.
As a conv~yor, the array processor executes arithmetic operationa according to the
expression z= y� vx. .Vector operations, spectra correlationa, complex multipli-
cation and fast Fourier transforms are performed in the array processor. Its basic
cycle is 155 ns and it is fitted in a standard Unified System rack.
~
The array processor is used by a set of acceas method programs that expand the
capabilities of YeS OS.
The high throughput of the computer complex made up of the YeS-1045 and the array
processor permits a substantial increase in the efficiency of solving problems on
_ pattern recognition, processing of geophysical data, etc.
COPYRIGHT: Izdatel'stvo "Finansy i statistika", 1981
8545
CSO: 8144/0138
!~2
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2407/02/09: CIA-RDP82-00850R000500440033-6
" FOR OFFICIAL USE ONLY
VIDEOTON FAMILY OF INTELLIGENT ALPHANUMERIC TERMINALS. -
Moscow VXCHISLITEL'NAYA TEKHNIKA SOTSIALISTI~IESICIKH STRAN .in Ruasian No 9, 1'981
(signed to press 21 May 81) pp 41-47
~ L~'ticle by L. Nylas, G. Hudoba and J. Bornemi8sza, all engineers from the
Hungarian People's Republic~ ~
[Text] Producta from the Videoton computer hard~tare plant, which has a decade of
e~hlrience in manufacturing computer hardware, are in great demand among both
- Ht~t~~a~ian and foreign specialists. The popularity and wide diatribution of Video-
t~ ~1E~?icea is dus not only to their technical parameters, but also to high quality
an~ fine eathetic and hwnan engineering characteris~ics. An example of the resulte .
achieved by Videoton is the fantily of intelligent alp~anumeric terminals based on .
microproceseors.
Microprocessors were deveioped ~o aimplify data processing and control hardware,
increase its reliability and reduce cost while raistng functional flexibility at
the same time. Microproceaeor throughput approach~s that of earlq small canputers.
However, to use them as computers requires expanding main atorage, connecting
extermal storage units, providing facilities for interruption processing, etc.
Microproceasor syatem software development is proceeding in two directions; first,
microprocessor sof tware is being developed on so-called HOST machines, and auxili-
arq software for development, euch as ~,ssembler and Simulator, is run on this ma-
chine; second, microprocessors are being made ca~a~ble of software devlopment.
Videoton chose the latter path.
In 1974, micropr~rceasora began to be used at the .plant in data tranamission and
processing systems. The VTS 56140 unit with mic~o~procesaor control dependent on a
built-.in program was developed at that time and used as a synchronous or asynchro-
nous terminal. From a functional point of view.,. Lhe most advanced model in the
VTS 56100 terminal li~e is the office vidgo tarminal with two external storage cas-
~settes. It can be programmed by the user, used 'in large systems, provides extensive
offline functiona for data baBe maintenance, atoring and editing; because of this,
this unit is recommended primarily 'for use in the-noninduetrial sphere.
3n 1976, development began on two, easentially new display families which then wen't
into series productioa. The develo.pment was based on the experience gained in the
- development, production and use of the well-kno~rn V2 340 dieplay (YeS-7168 and
SM-1206), as well as the VTS 56100 microprocessor ,family of video terminals, and
43
FOR OFF[CIAL US~ ~NLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
FOR OFFICIAL USE ONLY
with regard to the requirements and capabilities apecified by the Unified System
and System of Small Computers programs.
The one family is built with basic devices with small power consumption, made with
highly reliable LSI circuits and they provide great functional flexibility and
broad capabilities. The early deviceai~r~thie family are the SM-7219 terminals
(which passed international tests within the framework of the System of Small Com-
puters in 1977), the YeS-7168M terminals and improved YeS-7168 terminals.
Devlcea in the other family are terminal stationa: intelligent alphanumeric video
terminals with a set of external devices, easily programmable by the user in
Assembler or a high-level language. T~vo devices in this family have been approved
at the international level: the SM-7401 which has a built-in minicassette storage
unit (which passed joint testa within the framework of the System of Small Compu-
ters in 1978), and the VT 20 equipped with large disk storage. These devices meet
world standarda in reliability and functional capabilities in their catego~y. ,
A measure of the intelligence of these programmable termianls based on microproces-
sors is the aggregate of these capabilities:
expanded random access storage that enablea loading and running of various user
programs;
connection of high-capacity external storage units, forexemple, NML [tape drives],
NGI~ [floppy disk units] and NrID [disk] with permanent or replaceable diska that
support processing of data filea;
connection of paper-tape IO devices to maintain compatibility with older eystema;
connection of serial or parallel pr inters; and
connec tion of various parallel and aerial data tranamission channels.
The capability of connecting external devices to an intelligent alphanumeric termi-
nal is shown in the block diagram (f ig. 1). Naturally, all connection capabilities
cannot be used at the same time in one device; besides arrangement problems, this
would cause a shortage in the time resou'rce for the central unit. However, through
suitable selection, one can build a configuration most auited to a apecific applica-
tion. ~ ~
Providing the circuit capabilities liated above and others still does not make a
machine intelligent: Without an eff icient software system, all this wo~ld remain
just a possibility.
The basic unit in a Videoton intelligent alphanwneric terminal is a microprocessor
minicomputer structurally integrated with an image generation ayetem. Let us list
the components of the block diagram of the device (see fig. 1) and their character-
istics.
Central processor: byte processing of data.
instruction execution time of 2-9 micr.oseconds
seven internal registers
interruption processing
three internal timers
storage, exps:~3able to 64K bytes.
. ~
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/42/09: CIA-RDP82-00850R000500040033-6
FOR OFFICIAL USE ONLY
~ '
. ,
([NSf70flMbltl~ ~ ~ C q '
Qtut/xpOxxe~u ~v aa= ~
Moar~ � .
_ u a, a e a o
b g
O ~y 4 V
/ 8 e
MrrKponpoyeccop 0
epamuexaA ` ~uN=pQHMb/f1 ~ (3.) 5nox ~ i E o o n~'
Micro rOC@8 naMAmb acuHxpoxxb~u COApAXlNItA rN ~~e C
p StOY'8gE uHmtpptuC xadiuanrvOb~
~ ~4 . 5 (6 )
(7) QeyxemopoNxuri eeoa-eaeoa u ~ruHa xaemM~
y~IpoeneNr~~ 6~roK conpa~ceNUa 6noK conpwxtHUa ~Nxmpp~Rtiic NMA Nxn~epqve:ic yK~ep~e~ic
u~o6vaxeHUta Al{I1Y HMQ ?ro nurru-xccctmt QM Qeoda- Ie~eoda c
nea~onexmer
(13 )
.
� Printer O
i .'~+mop AU~Y �
~~:~r. ~
' ~ ~~M~ MIlNLf ~KCCCBMNO/K O
Maxc.4r4+5 M6auM HMA ~18~
4X?, 5*2,5 M6arirtc 2 x l00 K6ari?n ~Q M Beod � eereoa C
~17~ nep~onenmb~
Fig. 1. Consolidated block~diagram of family o~ intelligent aphanumeric video
terminals .
Key:
1. synchronous/asynchronous modem ~ 11. minicassette tape interface
2. synchronous/asynchrnnous interface 12. 'GDM [floppy disk] interfac~ .
3. keyboard interface unit 13. papertape IO interface
4. 20V 20 microamp interface 14. CRT monitor
5. cluster control unit interface 15. diek: maximtna of 4 x 4+ SM bytes,
6. parallel interface 4 x 2.5 + 2.SM bytes
7. two-way IO and storage bus' 16. minicassette tape storage,
8. image control 2 x 100K bytes
9. printer interface 17. floppy disks
10. disk~ interface 18. papeztape IO
Display: 10 x 12 dot matrix
2 x 64 alphanumeric charact~ra
64 spec~al symbols, #or example drawing of li.nes
Gre+ek letters etc.
ecreen capacity 25 x 80 or 24 x 80 characters .
~ f ield and line image orgattization.
~5
FOR 'OFFiCIAL USE ONL'Y ~
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
FOR OFFICIAL USE ONLY
Keyboard,: atandard typewriter atyle arrangement
separate numeric keypad for data processing problems
~ ~ . function keye
upper and lower case
autaaatic repeat of characters ,
sudible aignals.
Data transmission line: aynchronous or asynchronous
interface according to reconnnendation 24 of the IrIICKTT
_ ~ [International Telephone and Telegraph Consultative
Committee] or a current loop of 20V 20mA (S2, IRPS)
cluater control unit on aeynchronous line
~ maximum rate withs erial data transmisaion is 9600 baud
~ with the capability of progremming
parallel interface
- pr~granunable procesa for data transmisaion and error
protection.
IO devices: parallel or serial matrix printer
card reader
card puncher
External magnetic .
storage units: two minicasaettes with capacity of 2 x 100K bytes
two floppy diaks with capacity of 2 x 250K bytes
replaceable disks with capacity of 5+ SM bytes or
2.5 + 2.SM bytes.
External magnetic storage units can be optionally connected to the basic device.
Operating System. The intelligent alphanumeric video terminal (PAV) and the VT 20
- unit can be used first of all as standalone units easily programmable by the user.
The difference between them is that the PAV hae two built-in minicassette tape
storage unite, while the VT 20 operatea with~lOM bytes of disk atorage. Therefore,
the operating systems for these devices differ somewhat because the VT 20 must make
efficient use of the advantages and capabilitiee offerred by disks. The general
scheme of the software syatem (SPO) ia shown in fig. 2. A more detailed general
plan of the system is ehown in fig. 3.
The basic software for the programmable devices has the following programa:
resident pragram. The programmable read-only memory (PPZU) [PROM] contains the
permanent program that asswnes control automatically when the device is switched on.
It performs the initial adjustm~nt and provides the capability of loading and start-
ing programs written ori magnetic tape or disk;
monitors. The device has three monitor programs. The first debugs programs and
corrects errors; the second has functiona of programming in the PROM; and ~he third
can control test programe. The programs permit loading and startup of programs of
various formats, changing storage contents, printing storage contents, output of
stor,age contents on data medium in loadable format, and funtions of programming the
PROM; ~ �
programming languages. The Assembler translator tranelatea programs written in
Assembler and indica tes syntactic errors. It works in two passes, praviding thereby
use of postdefinition labels; ,
46
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500044433-6
' FOR OFFICIAL USE ONLY
_ , npo: paNNa - .
1� pptaxu~an?Op
- 06 ayrcrriaroutue 3~ a~wxu R?cv~aaxr~ukrc-
~Z~AAOtAar~Iaor ttpotA~Wpolaxw? xuI r.potAel?~?x
Fig. 2. General echeme of software system
Key: . ~ .
1. organizer program 3. programming languages
2. service programe ' 4. p~eventive programs
. .
Assemb r Pedaxmrraoiaxue Loadiri ebug- . ~pow~e npo~oa~
Acct~6aep u ucnpa~nexue 3atpylxa . ~Onraad,ra .ve~ o5ceyxuraxu~ �
~czoexoto ,~anr~mra,~s g itlg ~ nvomuaaxmuKu
3 }~/npa~nexut (4 ) Cucmt.~ra
dacaniter ynPQantxus
eeodo~r - ~ertodo.~r
Ma~Hw~,x�W Na- Awlrurr ~ttpfdawrt
~rortwnreas Ma .Auny, aaxxars
xotcernaz (ducxarl ' 6
Fig. 3. General plan of system
, Key : ~
1. editing and cerrection of source 4. IO control system ~
� material 5. casaette (disk) atorage
2. other aervice and preventive 6. data transmission line
programa
' 3. display control
$asic interpreter provides the capability oF running programs written in Basic.
The interRreter uses optional circuit capabilities of the device. Also, the
inatructibn set is expanded by operations on operations on character strings;
editing and correction of source material is used to create. alphanumeric files
that are recorded on magnetic tape or disk by means of the keyboard or display
scre~n and to change aource alphanumeric filea;
~0 Control. The VT 20 operating system with time sharing enables execution of
47
. ' FOR; OFFICIAL USE ONL'Y ~
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007142/09: CIA-RDP82-40854R040500040033-6
a�va~ vrr~a..iHL uJG V1VLY
of logical control of external devices;
preventive programs. These include the test programs that check all device units
and the test monitor that controls them;
other programa. The device has numerous programa used to solve problems often
encountered in practice. The most important of them is the data file~management
system on disk that was devploped for office systema.
The efficient operating syatem, numeroua user programs and flexible achematic
- design, the intelligence of the device, enable use of these devices in all sectors
of economica, starting with devices for data transmisaion and processing to
desktop camputers.
Recom~aended Areas of Application. In quer~? applicationa, the device furnishes the
operator rapid accesa to data bank file$ kept on diak. The device can be uaed to .
update files and for inquiries in teh interactive mode.
In data writing applications, the video terminal replacea ~raditional units, for
exemple a card puncher, and can be used Co update source files and store data in
the onTine or offline mode, making use of the advantages offered by the display and
external magentic storage units.
COPYI.:.~HT: Izdatel'stvo "Finansy i statiatika", 1981
8545
CSO: 8144/0138
48
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
FOR OFFICIAL USE ONLY
YES-2335 ARRAY PROCESSOR FOR YES-1035 COI~UTER SYSTEM
Moacoa VYCHIS'LITEL' NAYA TEKHNIKA SOTSIALISTICIi~3KII~Ii STRAN ia Russian No 9, ~9:81
(signed to press 21 May 81) pp 47-50
[Article by G. P. Nikolov and V. D. Lazarov, candidates of engineering science,
and P..P. Daskalov, Y. V. Ivanova and K. D. Kirov, scientific aseociates, all
from the People's Republic of Bulgaria]
[T~~ct? Constantly growing ia the number of users o� computer hardware for perform-
irig, ;apid and eff icient digital processing of large files of data represented in
- tb~ ~orm of arrays or vectora. Processing of such data files is required in seismic
e~~oration (geophysics, searching for oil, gas and mineral resourcea); in studies
of turbulence (meteorology, oceanology, aerodynainics); in radiolocation and proces-
sing of results of photogra~hy for enhancement~and processi;ig of images; in space
exploration (telemetry, trajectory control); in research in nuclear physics.; in
matrix arithmetic; in mathematical statistics; ~or solving differential equations
and digital processing of signals.
However, modern general-purpose computers, despite the continuous increase in speed,
are not capable of ineeting the requirementa of throughput and system computing
capacity for these types of problems. Execution of repeated arithmetic operationa
on elements of large f iles of data is practically unrealizable. Thus, it is
necessary t,o develop apecialized computer facilities to solve these problems.
There~are several ways of developing efficient arr~y computers. One of them, in
our opinion the most successful at the current e~age of development fo the Unified
System of Computers, has been implemented in the YeS-2335 array processor (I~). It
was shown in 1979 at the exposition entitled "Facflities of the Unified System of
Computers and the Small System of Computers and their Application." Let us consi-
der some characteristics of thia procesaor. .
Zhe YeS-2335 model-dependent specialized peripheral proceasor performs high-spe:ed
calculations within the YeS-1035 computer system. Proceasing in the array processor
is parallel and independent of that in the YeS-2635 CPU.
The YeS-2335 connects to the YeS-~i635 CPU through a specialized adapter that re-
places a selector channel in the CPU. The adapter and array procesaor are linked
by a specialized IO interface.
. ~#~s;
. 49
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2407/02/09: CIA-RDP82-00850R000500440033-6
rvn vrr~~.~r?i, u~C, UIVLY
From the system aspect, the array processor and adapter can be considered a complex
IO unit combining the functiona of a channel, control unit and external device.
This complex is controlled by using the same syatem of control worda used in the
Unified System IO system (IO instructions, KSK [expansion unknown], USU [~anit con-
trol word], SSK [channel status word]) and the IO interruption mechanism.
. Data processed in the array procesaor are repreaented in one of the following for-
mats used in the Unified System of Computers: fixed point, direct code, short for-
mat; fixed point, two's complement code, short format; floating point, short format.
Array processor operations can be divided into the following groups:
a) vector move operations: data are moved from one area of main storage to another
with execution, if indicated, of format translation (from fixed to floating point,
or vice versa);
b) matrix arithmetic operationa: scalar multiplication, element-by-element vector
multiplication, element-by-element sum of vectors, sum of vector elements, partial
matrix multiplication, sum of sq,uares of elements and square of matrix with sign; ,
c) matrix scanning (to determine value and position of minimum or ma.ximum element);
d) complex multiplication;
e) solvfng difference equations;
f) signal processing operations: convolution of addition, convolution of multipli-
cati~n, fast Fourier transform;
g) quadratic interpolatfon.
The array processor has the following assemblies and unita:
unit for communication and fetching of operands; it effecte communication with the
adapter and YeS-2635 CPU, processes addresses of operands and control words and
~ fetches operands;
arithmetic unit; it executes conevyor arithmetic processing of floating point
numbers;
buffer storage; it holds operanda, intermediate and final results'and performs the
functions of a buffer between the inetrface and arithmetic unit; ~
control storage; it holds microprograms for the operation of the communication and
arithmetic unita;
microprogram control unit.
The basic parameters of the YeS-2335 are:
machine cycle 200 ms
throughnut 5 million operations/sec (multiplicatian
and addition)
throughput of interface between adapter
and array processor 3.SM bytes/sec
throughput of arithmetic unit 20M bytes/sec
buffer size 2 x 32 words
control storage size 2K words (of 96 bita each)
50
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500044433-6
FOR OFFICIAL USE ONLY
r~aximum operand size 64,000 elements
element base TT-1.2 and 3
~tructure one cabinet with two frames
power consumption 1 kW
The basic aof tware for the YeS-2335 array processor consists of inethods of accesa
(MDI~P) and the resident module. It is implemented in the Assembler language with
use of ~he capabilities of the macrolanguage and .conditional asaembly gtatements.
The access method ia activated by a user program in the ~ORTRAN, Assembler or PL/1
languages by using the CALL statement having the ~pecifipd format. The resident
module enables proceasing of SVC interruptions with a code specifically for the
array processor.
The access method has the following parts: cont~ol phase, syntactic analyzer,
channel program deaigner, disp~tcher for queue of orriers [zakaz], issuing of ines-
= sages and processing of interruptions.
In addition to the basic, the array processor sof twa�:e also~includes a eignal pro-
- cessing program system. It is a package of subroutines that are used in the analy-
sia of varioua types of signals in digital form. The package contains subroutines
t~~t syntheisize filters by specified criteria for their characteristics, and sub-
rQ~}x~.t~es that help evaluate the efficiency of th~ filtera,
The lack of sufficient experience and the accept~d technique of evaluating through-
put of systems with the use af array processors prevent determining the eff iciency
of their applicatian and fixing the class of problems in which this application is
most expedient. To derive approximate estimatea of the increase in system through-
put, experiments were conducted in which the samQ specially selected examplea of
data processing were executed on a YeS-1035 computer with and without use of the
YeS-2335 array processor. They confirmed the preliminary estimates that by using
the array processor, throughput is raised by one-two orders.
Results of one experiment are as follows. An average-size seismfc route was
divided into 2000 quanta. The array processor w,as used for certain basic procea-
sings such as deconvolution before and af ter summing, �iltration before and afrer
- summing, and use of a fan-shaped filter for 10 Rercent of the data. In thfs case,
proce~sing time�by the CPU decreased by 65.6 s for the route. Annual process,ing
of 300,000 24-channel recordinga by usir.g the array processor takes about 4,Q00
hours. Consequently, 135,Od0 hours are required for a system without the array ~
processor, i.e. throughput is increased 34-fold.
COPYRIGHT: Izdatel'stvo "Finansy i statistika", 1981
8545 � ~
CSO: $144/0138 ~
~ 51
FOR OFFICIAL USE 4NLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500040033-6
rva~ vrr~~,i,~i, u,r. UNLY
SYSTEM FOR DESIGN OF APPLICATION PROGRAM PACKAGES FOR CALENDAR-SUBJECT PLANNING
AND MANAGEMENT FOR UNIFIED SYSTEM COI~UTERS
Moscow VYCIiISLITEL'NAYA TEKHNIKA SOTSIALISTICHESKIKH STRAN in Russian No 9, 1981
- (signed to presa 21 May 81) pp 51-57
(Article by V. A. Matulis, candidate of physical and mathematical science, and
A. A. Chaplinskas, engineer, both from the USSR]
[Text] The "Vil'nyus" system for deaign of application program packagea (PPP) f
calendar-subject planning and management of development work is based on systems of
network models for a multilevel automated system for planning and management of a
machine-building industry [1]. The problema s~lved within this automated syatem
form a unified complex of problems; therefore, they must be correlated with each
other. Correlation is provided by the informational, organizational-methodological
and program-mathematical compatibility of the probleme. Also provided for is the
use of uniform indicators, a unified eystem of classification and coding of techni-
cal and economic information and unification of reporta output.. Taken into ac-
count are the presence of information exchange between the individual complexes of
_ problems and the intercorrelation of problems be~ng developed (etructure of plans,
iridicators, objects and procesaes of planning, resources, organizational structure
and others). The system is based on principles close to those used in program
generators for proceas control [2].
The "Vil'nyus" sysCem is parametric [3J; its nucleus unites procedures that imple- ~
- ffient input languages and data base management for the given class. In constructing ~
PPP, in the system input language, the user describea the package input language,
the data base stru~cture and other parameters.
The basic concepts of the "Vil'nyus" system and ita implementation methods are
discussed in more detail in [4, 5].�
Requirements for Packages in the Given Class. Studies have shown that PPP for
calendar-subject planning must include~: ~
a nonprocedural input language that permits formulating the aims of the design in
the form of a specification for deriving a certain aet of output documents;
_ facilities for automatic planning and implementation of the plan for solving
problema; .
facilities for automatic adjustmefit of the plan for solving problema as a function
of results of intermediate calculations;
facilities for input of i,nttial data that provide the capability of punching data
52 .
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007142/09: CIA-RDP82-40854R040500040033-6
FOR OFFICIAL USE ONLY
directly from user ~cqrking documents;
data base indicators that permit determitting the current status of each of the
data segments; and
- facil~,tiea for building a temporary Working model of the planned object that
allo.wa keeping in the computer undeleted complete information on the plan through-
out the entire plan period and permits making test calculationa to determine the
consequences of some plan adjuatment before its off icial approval or multivariant .
calculatione in drafting a plan,for a new period.
Ttie PPP must provide the capab.ility:
of several related PPP using a common data base; problem modulea must be indepen-
dent of the data base structure;
of segmenting data in a data base (by subjects, aubdivisions, etc.) and arranging
datya wfthin a segment in the form of a heirarchical sequence of records;
of specifying a model of the subject in the form of a network that has several
initial and objective events and permits directive reatrictions on the time periods
for completion of any specified events;
of including in the model~linear operations or ~,pecifying them in the form of a
linear graph; and
of automatic adjustment to the computer configutatfon and parameters that reflect
the specific nature of specific user problems.
Ia~_~ition to what is liated above, there are a number of other aspects that '
af�g~t mainly the content of the library of problem modulea and its functionai,
capabilities. They are not considered in this work since they are not esaential.
in defining the basic concepts of the syetem part of a PPP.
Syetem Composition and Structure. The system includes the monitor, the set of
problem-oriented subroutines and macro inetructions, data base management programa,
library of system tables, library of problem modules and the language for
describing the structure of the PPP for generation facil.itiee.
Internal control of tasks and ,jobs is performed b.y the monitor. ~
The monitor, one of the main components of the nucleus, performs decoding of tasks,
forms the plan for solving a problem, checks the authority of the user, implementa
the plan for solving a problem, checks the informational provisioning of the com-
puting proceas, adjusts the plan for~solving a problem and managea the listing of
its solution. '
The "Vil'nyus" system has a modular structure of three levels. Subroutines and
macro inatructions make up the lowest level of the problem modules. In essence, .
they form a eertain macro language used to program modules in the next level, the
macro blocka. Macro blocks are problem modulee defined in terms of algorithms for
solving problems in the subject field in queation. Only modules of the highest :
level can be described in terms of the subject f ield itself; these are proceasors ~
that are organized sets of macro blocks.
The acheme of the potentially permitted calculations of any PPP for the claes in
question is represented in the form of a sequence, ordered with respect to the time
axis, of segments that are describ~u by oriented acyclic graphs with one terminal
and or:e initisl vertex. Some procesaor is set to correspond to each segment, and
53
~
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007142/09: CIA-RDP82-40854R040500040033-6
N'UK UH'FICIAL USE ONLY
eome macro block ia compared with each of the vertices (except the initial and
terminal) of the oriented graph. The ares define the potentially permiasible
order of execution of the macro blocks. The procesaors are described by tables
that specify the topology of the oriented graph, establ~sh the method of mapping
the set of its vertices in the problem module library and determine the input ~rid
output processor data flow.
1he data base management programs provide for loading and updating of a data base,
creating a working model for test calculations, internal translation of data and
output of results [6].
Problem modules do not have direct access to a data base. .They operate with work-
ing sets of processor data formed by the facilitiea for internal translation of
the data prior to execution of the procesa. After its execution, the results of
the calculations are moved from the work aets to the data base by ueing the same
facilities. Formation of work seta and recording of results are effected upon
manitor requests issued in accordance with the plan for solving the problem and
the deacriptioq of the proceseor data flow. The description of the data base
structure and~~'vork set structure is kept in the form of special system tables,
which are used also in printing out the resulte, conaidered ae one of the typea of
= internal tranalation of data. Tablea describing patterns of punching the aource
documents are optionally used in data input and update. .
The problem of loading and updating a data base is considerably more complicated
than .that ot internal translation of data. This is due in particular to the fact
that the correspondence between source dociunents and data structurea being gener-
ated cannot be established by uaing apriori descriptions, but must be defined by
software. Input and update programs are organized as an independent proceseor
included in any package being deaigned. The tasks for this rpocessor also include
file management, operation with unconfirmed updates and formation of a working
model used to conduct test calculations.
Operation of the monitor and data base management progarms~ is controlled by system
tables; the main ones are the catalog of macro blocks, catalog of processors,
requisite description table, file description table, punch pattern description
table, table of inessagea and input language description table.
Several interacting PPP cney be built on the same copy of the nucleus. Each of them
may have its own input language, echeme of potentially perraissible calculations and
data structures and conduct a protocol for solving a problem in its own terminology.
Consequently, each of the packages has ita own corresponding set of system tables.
All sets are organized into a common library. The packages operate by using the
common library of problem modules. However, each package has its own procedure
for starting the monitor and operates under control of the set of system tables
_ determined by this procedure.
Formation of system tables is one of the main stages in designing PPP by using the
"Vil'nyus" system. This atage is completely automated. A special language has
been created for this purpose [4, 6] that permits the develoger to describe fn a
convenient way the package input language, data base, patterns for punching the
source documents, processors, texts of inessages to be issued and the calendar
(number of work days in a week, method of determining work Saturdays, units of
51~
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
- - ~ FOR OFFICIAL USE ONLY
~easurement of the time interval, etc.). The system tables are generated accord-
ing to these descriptions by using sof tware facilities included in the nucleus. �
To describe the package atructure, one has to know well the principles for design-
ing the structure and functioning of a PPP, upon which the "Vil'zyus" system is
baeed. No knowledge of the internal design of the system itself is required.
The generation f acilities are intended for adjusting a syatem to, a specific compu-
ter configuration and specific nature of user problema. For this purpose, the
apparatus of cataloged procedurea and facilities of the macro assembler are used.
Computer parameters~ and problem parameters are described by using the generation
form. A special program, operating directly under control of YeS OS, examines the
procedures and macro instructions, adjusts them according to the values of th~,
parameters specif ied in the form, performs translation of the system and createa
the data sets needed. A working version of the syatem is generated as a rest~lt.-
Problem Modulea. To cover the functions of the problem area in question, in the
authors' view, the library of problem modulea qnust contain modules for calculation
- of time characteristics of the n.etwork and load of the executives and modules for
balancing the plan by time, cost and other indicators, dfstribution of resources,
formation of plan indicators and expeditious control.
- Deve.~opment of modules of these classes for the.machine building aector is a com- �
p~~~ ~nd laborious task. Standard modules only for calculation of time charac-
t4~i}~tics have been included in the problem module library of the "Vil'nyus" sys-
tem. They are oriented to models of subjects with a capacity up to 16,000 opera-
tions; maximum nu.mber of subjects ia 999. Linear operations are also allowed in
the models along with network. A network may have an arbitrary number of initial
and objective events. Directive reatrictione are permirted on the term for can-
pletion of any evente in the network. The type of networks is "operation-arc_."
The set of problem modules permits: ~
combining individual fragments (subnetworks) into a cortmaon network;
performing analyeis of a network as a whole or by fragementa to detect structural
errors and circuita; .
calculating time characteristics by uaing a mixed linear-network model;
removing executed operations and automatically determining terms of initial events
of an adjusted network according to report information;
isolating from a network a subnetwork that enables completion of some set of
specified (check) events and calculating the time characteristics of this subnetwork;
forming according to a specified set of indicators in a specified profile output
docunents on time characteristics of operations, time characteristics of events
and executed operations.
Dates of initial or objective ev~nts in the network may be used as the source in
calcnlating time characteristica. Machine-orie~ted algorithms given in [1~ were
uaed in designing the modules. ~ .
Designing Packagea. To design a package by using the "Vil'nyus" system, it ia
necessary to generate a system that takes into account the computer configuration
and the speciffed values of the problem parameters of the problems to be solved,
to write a procedure, to start the monitor, to describe the package structure in the
~'Vil'nyus" system language, and to expaad the library of problem modules with the
required macro blocks.
55
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500040033-6
Nutc uhhll;lA1. US~ UNLY
In the case of expanding the aubject area with new concepte, it is necessary to ~
write additional subroutines for the meaning translation of data. Subroutines are
called during entry of data which describe the new concepts. To include the sub-
routines in the package, they must be described in the system input language.
System Use Experience. Let us cite two exemples of using the system as an illus-
tration of system capabilities.
Example 1. A program package to generate on a computer a five-year subject nlan
for development work in the sector. Zhei package is designed to formulate five-
year plans for development by the sector of technical complexes under the condi-
tions of the restriction on resources available to the organizations and enter-
prises and the amounts of state budget appropriations allocated to the sector as a
whole for development work. Taken into account in the procesa are the relative im-
portance of the development wcrk being undertaken and the minimum permissible and
maximum possible rates of development of the individual technical complexes, their
parts and systema.
Included in the complex of problems for five-year subject planning of developmQnt
work in the sector are:
calculation of~technical and economic indicatora and generation of the initial
alternative for the five-year subject plan for sector development work;
- optimization of the volume indicators and generation of an optimized five-year plan
for sector development work;
calculation of the development ~ffort workload on organizations and enterprises;
generation of excerpts from the aubject development plan by main directiona, orga-
nizations and enterprises in the sector, and of subject plans-orders for
co-performing ministries.
The basis of the mathematical model of the planned object is the system of network
models of a special type with the specified structure of developments. The struc-
ture reflects on the one hand the development stages, and on the other, the
component elements of the technical complex being developed.
Example 2. A program package to generate a development subject plan for subdi-
visfons of a KB (NII) [design bureau (scientific rea~arch institute)] for a quarter.
This package is designed to generate quarterly plans that provide for performing
work on a aubject by a subdiviaion, united by the feature of the direction of
research or functioning of items. The end product of any the~ne is a prctotype.
The complex of problems for quarterly subject planning for a KB(NII) includes: ~
calculation of Che technical and economic indicators and generation of a draf t
plan for the quarter being planned;
generation of a report for the last quarter;
generation of a quarterly plan taking into account changes made to the drafC by
management and information reports;
accumulation of information reporte during the quarter and generation of plan
supplements at established frequencies.
The basis ~f the mathematical model for the planned ob,ject is the system of linear-
network models that def ines L'ne structure of the development, the atructure of the .
prototype being developed and the structure of the organizational and financial
links between the separate performing agencies.
_ 56 ,
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2407/02/09: CIA-RDP82-00850R000500440033-6
FOR OFFIC[AL USE ONLY
Bott~ packages were siesigned with the aid of the "Vil'nyus" system. In the first
package, two of the four proceaeora, and in the second, two of the three proces-
sors were designed completely~on the basis of the exiating problem module library.
They effect entry of the initial data and calculation of the time characteristics
for the networks. Also developed for the f ive-year palnning package were modulea
that distribute resources and generate output documenta, and for the quarterly
planning package, modules that perform calculation of the plan indicators and
generate output documents. In both cases, the syetem portion was completely
built with "Vil'nyus" system facilitiea.
System Technical Characteristica. The "Vil'nyus" system was developed for the
Unified System of Computers; it operates under control of the YeS OS operating
. system with capabilities of at least that of version 4.1.MVT without subtasks.
For system operation, 100K bytes of main storage a~nd about 7M bytes of disk .
storage are required. Storage needed for executipn o� the PPP designed ie a
function of the featurea of the package itself. The resident part of the syatem
in the process requires 4K bytes of main storage. The aystem nucleus is
programmed in Aasembler.
Proapecte for System Development. System uae has ahown that the design stage
requires the greatest efforts. Design of a system according to a structure de-
sig~ed in advance and modification of it are then performed relatively easily,
~~~hQUgh inaufficient universality of some tools has been noted, in particular
t~~ facilitie~ for data base loading and updata.
Further development of the "Vil'nyus" system is planned in the following directions:
development of new problem modules for more complete funetional coverage of the
subject area;
expansion of the capabilities of existing modulea to offer the user a larger
spectrum of capabilities for.specification of networks (includi.ng introduction of
the "operation-arc" type network, use of analogs, standards, etc.);
expansion of thQ problem-oriented macro facilities by f acilities based on the
concept of abstract types of data;
development of specialized facilities for generation of output documents;
development of interactive facilities for generating and optimizing network
models and step-by-step improvment of plans.
BIBLIOGRAPHY
1. Tamm, B and Tyugu, E., "Developtnent of Problem-Oriented Software,"
KIBERNETIKA, No 4, 1975, pp 76-85.
2. hfusstopf,. G.; Orlowski, H. and Tamm, B., "Program Generators for Procese Con-
trol Application," The 2nd IFAC/IFIP Sympoaium on Software for Computer Control,
SOCOCO ~79, Preprints, Vol 1, General Computing Center of Czechoslovak A~cademy
of Sciences, Prague, IP-III, pp 1-~,2.
3. Red'ko, V and Yuzhenko, Ye., "Algorithmic Languages and Translating Systems,"
KIBERNETIKA, No 5, 1961, pp 87-91.
4. Matulis, V, and Chaplinskas, A., "Structure of Program Package f.or Calendar
5ubject Planning System," in "Struktura i organ~izatsiya paketcv~ prikladnykh
program~n" [Structure and Organizations of Applicatio~ Program ~ackages],
Tbilisi, 1976~ pp 117-119.
57
FOR OFFICIAL USE ONLY '
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2407/02/09: CIA-RDP82-00850R000500440033-6
N'UK Ur'N'1l;lAL USE ONLY
5. Vaychyulis, B.; Matulie, V.; Sinyuvene, R. and Chaplinskas, A., "Some Problema
in Designing Application Program Packages," in "Avtomatizatsiya protsesaov
planirovaniya i upravleniya" [Automatfon of Planning and Management Processes],
Vil~nyus, No 3, 1976, pp 9-47.
6. Tonkikh, V.; Chaplinskas, A.; and Yuozevichute, G., ~~Organization of Data in
the 'Vil'nyus' Syatem," in "Automation of Planning and Management Procesaes,
No 6, 1918, pp 11-26, Vil'nyus.
7. Vaychyulis, B. and Matulis, V., '~Coding and Machine Calculation of Time
Characteriatics of Large Networks," NAUCH. TR. VYSSHIKH UCHEBNYKH ZAVEDENIY
LITOVSKOY S$R. AVTOMATKIA I VYCHISLITEL'NAYA TEKHNIKA, III, 1970, pp 225-234.
COPYRIGHT: Izdatel'stv~ "Finansy i atatiatika", 19$1
8545
CSO: 8144/0138
58
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
FOR OFFIC[AL USE ONLY
~DIGITAL SIMULATION OF CONTINUOUS PROCESSES ON UNIFIED SYSTEM C01~'.UTERS
MoscoW�VYCHISLITEL'NAYA TEKHNIKA SOTSIALISTICHESI~IKH�STRAN in:,Russian Nc 9, 1981
(signed to press 21 May 81) pp 58-65
[Article by Ya. Yazhembek, Master of Science, Po,liati People~s Republic]
[Text] Irbdern programming languages and continuoua.proceae simulatiAn systems.are.
used exteneively~for scientific and industrial ;eaearch. Digital aimulation of, the
~ttu,al procesaes yields more accur~te results than analog. Therefore, many modern
~n~al-purpose computers are equipped with a special aimulation apparatus. Until
~.~.~ly, Unif ied System computers had no apecial software for simulation, and the
onS~ capability was implementation of simulation~algorithms in a high-level lang-
uage (FORTRAN, PL/1). H~owever,, this requires a good knowledge of these languages~
- and much experience.in prog~amming.
The DIANA programming language along with t~l~: we~1L-known CSI~ language fills thia
gap in Unified Syatem software operating under eontrol of DOS or OS. y
General Deecription of DIANA. DIANA is a problem-oriented, language providing the
capability of digital simulation on Unif ied System computers.with the YeS`DOS
operating syetem. It makes it possible to aol~e,problems associated with simulation
of arbitrary sqstems described by d~ifferential equations or an analog block-diagraa?.
A general diagram of the functioning of the DIANA.language and.s.imulation eystem
is shown in f ig. 1.
A DIANA program is block-oriented, i.e. developed on the basis of a block diagram
that repre'sents a model of a real system. The. programmer, by assigning to each '
block in the diagram a.stetement that defines~ the action of this block (integration,
delay) and a name that representa:the output signa~ of this block, can describe�.the
structure of a system by using so-called atructural inetructions. For example.,
the ins.truction Y=(INTEG X, X0.) executee integration of th~e equation
dx = x(t) with the~ initial conditions of x(t~ Q:) x~.
The aeC of structural instructions is used as the be,sis for defining the aequence
of calculations according to. the correaponding algor,iChm. A DIANA program is made
u,p of several parts separated from, each other b~ service records. Ttre eimulation
, algori~hm is placed betwee~ the PROGRAM and END records.arld can consist of three
ereas: initial, dynamic, terminal.The initial a~rea f.s a series of FORTRAN instruc-
tions bounded by the INITIAL and END records. Th~~inetructions in this area are
59
FOR OFF'ICIAL U6E ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2407102/09: CIA-RDP82-00850R000500440033-6
cvn vrr~~,ir~a, v~C. V1~LY '
executed prior to the start of the simulation procesa and ar~ needed for calcula-
tiona aseociated with the definition of values of"parameters in the dynamic area,
initial conditiona and gain factors. ~
analo mathematical
AMa~?otosaA � Mamt~amuvt �
6aok-cre~a Ka~ ~rodtas �
block-diagr model
DIANA /lpotpQ.v.raxa '
program ~
. Aie~Ke DIANA
D/ANA
[s
nif ied ~
_ ~System Pe~yabmame~ gimulation
~ 38M EC cW~y~'pu~`~ reaults
- ~Computer ; ~
I ~
i
Fig. 1. DIANA system functional diagrem
The dynamic area is bounded by the DYNAMIC and END records; it is a series of
instructions that describe the simulation process. This area~contains the program
that executes ~alculations for one step of ~ntegration. At the start of the area,
the programmer can define his own atructural blocks by describing them as MACRO.
The terminal area is a seriea of instructiona bounded by the TERMINAL and END re-
cords; it is executed after completion of each computing cycle, i.e. a slice of
time, after which the program can report results through an IO device. In thia
area, the programmer can change the parameters for a multiple pass of the program
or testing of certain conditions.
FORTRAN instructions may also be placed before the initial area and af ter the
terminal area to have them executed before and after, reapectively, the simulation
process. Here is a general view of a DIANA program:
60
FOR OF~'ICIAL USE ONLl~
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/49: CIA-RDP82-00850R040500040033-6
FOR O~FICIAL USE ONLY
Nota.~ Zhe F1 area can also contain sub-
PROGRAM routinea in the form of SUBROUTINE~or
SUBROUTINE P1...
, FUNCTION. The IC1 record is a aet of
� instructions that def ines the initial
FI RETURN conditions. IS is a series of atructural
END instructions. IC2 defines the ~erminal
� ' conditions. F2 fa a series of instruc-
; ' tiona executed after completion of the
Apyree xoMaHAw ~oprpaaa aimulation prc~cess. GO is a directive
other FORTRAN instructions that ~ef ines the start of the simulation
' process.
INITIAL ~
1C1
END ~
DYNAMIC ~
MACRO
END .
iS
END ~
TER3lINAL
IC2
END
~ F2'
GO
STOP D1ANA
Structural Inatructions. The basic structural block uaed in DIANA programming is
the INTEG integrating block that has as the input variable the name of the signal
subject to integration and the initial conditio.na, and as the output variable.,
the name of the signal�after the integration operation. The structural instruction
y = INTEG (X, WP)
corresponds to the formula . ~
r .
Y(t)=WP-~-~ X (t)dt; ~(!=0)=WP. ~
~iql0
In this formula, t denotes the independent variable TIME. This structural state-
ment has no argumenta and denotes time~as an independent variable during the simu-
lation proceas. Associated with the INTEG statement is the ALGORITHM directive
that allows selecting the integration method: either Adams fourth-order method or
the Runge-Kutta fourth-order method. The Runge-Kutta method is standard..
In addition to the integration statement, DIANA includes the following functional
blocks:
DELAY delay of integration by a step
HYST hyatereais of people?)
DER differentiaticn b.lock
RST RS-trigger
61
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
NUR UFFlCIAL USE ONLY
BLOCK . ieolation block
RANDOM noise generator ~
DSPLACE dead zone (insensitivity)
LIMIT limiter .
. FG arbitrary function generator
AND operator for logical multiplication
NAND negation of logical nrultfplication
IOR operator for~logical sum ,
NOR negation of logical aum
NOT negation of signal
COI~ARE comparison of signals .
_ EOR non-equivalence
EQUIV equivalence .
QNTZ R , quantizer operator
SINE generator of harmonic oscillatione
INSW input ewitch .
FCNSW functional switch
RA1~ ramp function ~
STEP etep function generator
In addition to theae functional blocks, the DIANA language also permits all
~ standard FOR2RAN~language functions.
Z~anelation and Simulation Results. During translation, a DIANA program ia trans-
lated into a set of interrelated pro~cedures in the FORTRAN language that execute
the atructural instructions listed in the pragram. Then proceseing is performed by
the FORTRAN translator and the simulation process ie executed. Upon completion of
the latter, results are printed out in tabular or graphic form. The programmer
obtains a table of values of variables as a function of time or a graph of a certain
variable as a function of time. The phases of processing of the source prog;am,
starting with job definition to obtaining simulation resulta, are shown in fig, 2.
Let us take the following problem as an eacample. ~
Example. Let there be a system of chemical reactioas of the first order, the con-
stants of wl~icfi. are Kl and K2, respectively:
~ ~ " A KiBx:C
Let (AQ~ be the initial value of factor A at time t~0; let us assume (H~)= (C~~=Q.
Then the system of differential equations that represents.th.ese reactions takes the
_ form
_ dx a- KI�x7 dy a Kl �x- K2�y; � az a K2�~I~
~ where dt dt dt
x= (A)/(ao);
~ 9~ (B)l(Ao) ~ . .
~s (C)/fAo)
denote the relative values of factora A, S and C, respectively.
62 ~
FOR O*~'FICIAL USE ONLY ~
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED F~R RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
FOR OFFICIAL USE QN~.Y
.Rey:
Onptaeetx 1. prohlem definiti,on
,~adavtr `~~.1 2. source grogram
DIAN~ 3. DIANA ~ra~slator
I, FORTRAN traaslator
yy editor
NcsoBxaA (z , executive routine
npc:vax~ra 4. FORTRADT lihrary
5: sf~mulat~.on results
baMuomtxa
~ ~Popntpax .
~4 ~ ~
t
jTpaHGO,~ro/ OUNA
~lOQNi/lAiOpAlIfKQ ~ P~~ypblrtQlRb~
j mOA+nAaM CuMyAANtlu
~ Atdarrmop
~ ~ NtrtonNUttnexoa
~ n0atpa.wXa '
Fig. 2. Source program processi.ng phases
Analytic integratir~a of this system yields the salutinn
z(t)~exp (-Kl�t); ~
y(t)=sK2KK~ [exp(-Kl~t)-exP(~K2�~)~~
Z(~)=K2: Kl exp(-K1�t)-?-K2K~K1 exp(-K2�t):
- Tfi,e DIANA program for this example is given in fig. 3. Figure 4 ill.ustrates the
resul~ts of simulatiri~ this prohlem for factor A.
DIANA System Operating 'r:xper~ieace. The DIANA system for simulatinn of continuous
processes ~as developed and is. in use at the Institute of Control Systems in
Kato~rice ~Polish People's Republic~.; it is also ia use at the Institute of Mathe-
matical Machines in Warsa~. System operating experieace allows making some com-
parisons. The DIANA simulati~on system is similar to the CSI~/360 spstem. The time
for executing similar simulation algoritbms and the accuracy of results ohtained
are about ideatical. DIANA language fuactions cprrespond to CSMP language proper- ~
ties, hut th,e cTzoices in method of calculating th.e correspondiag functioas are
limited. Ti~us, DIANA offers t~o meth;ods of integratina, the Adams and the Ruage-
Kutta. In addition to these, CS1~ offers tbe methnds of Simpson, Milae, trapezoids
aad rectaagles. The maia advantage of the DIANA system is the capability af usiag
it ~ritF~. smaLl comgdter configurations ~ritii 128K bytea of storage (syatem si.ze is
ahout lOBK hytesZ. Morp than 128K lrytes of main storage and~OS are needed for the
CSr~ system. ~ ~
~ 63
FOR OFF[CIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
FOR OFFICIAL USE ONLY
' P11~ORAM
INITIAL
C1+SK1/llK2'SK1)
C2�'SK2/IJKY-SK1)
!ND
; DYNAMlC
%PN1M+-SNt~%
_ 1c�INTECIXPRIM;xOi
rOBIM=SKt�x-SK2�r
r=INIEGIYPIliM;rC~
2pq]M�SK2~Y
Z�INTECCiPRIM;LO)
Ros1�-Snt�iIME .
AV�ExPlR061) ~
ROSY�-SIt2�TIME
~ ROi3�EXP~K062) .
p084�1cP-p003
Fig. 3. DIANA simulation program ra=c~�ROe4
2P=ONE~C2~XP+CI~ROBd
END
ENa
TMINCG.O~
TMAX(5.01
- TSTEPCO.t)
NSTEPS(1)
ALG01121MM(RK)
OATACIIO~1.iY0~0.i10~0.rSK1~Z.O~SRZ~~.O) �
TI1~E(tHt,aox,zeNwrN1KI ?ROC[SU MOOELOMANIA////)
r~tnt(x,xv,r,rv,t,zp) '
PIlTPLOT(x~xv,r,rp~i,iP)
GO � ~
iTOP DIANA
11ME X M1NiNUM. 0.4!~Of1E�Oq Max1MUM. E.~OO000E i,~
o.a o.~oooooE o~
o.,ooooc o0 o.a,e~sae oo i-=-------------~...----------------------------------------��
o.zooooe o0 o.a~oixsa oo t------�----------------------------------------'
o.~ooooe o0 o.sree,re oo t---------------------------------------�
O.r0000E 00 0.4~9!!SE 00 1-"'-'---'----------�������
O.SOOOOf 00 0.301!!6E 00 1-------------------
0. s0000E 00 0.301200E 00
O.TOOOOE 00 0�2V6607E 00 1-----------'--~--~�
o.eooooE o0 o.zo~vo2e oo t----------------~
0.90000E 00 0�10l304E 00 1------------+ ,
0.10000E 01 p�t33310E 00 I----------~
0.11000E 01 0.110lOf8 00
0�12000E 01 0�f07Z1Sl~01 l-------� ,
~ 0�13000E Ot 0.142707E'Ot f-----�
p.1h000E 01 0.601129E-01 I----~ � ' ~
0.~lOOOE Ot 0.~~~1~lE�01 i---�
O.~OOOOE O1 ' 0.40~O~yE-0~ 1--�
O.t~o00E Ot 0.!!71ltE-0~ t--�
0.16000E Ot 0�=1~tlIE�01 I'~
9,~voooE ot 0.227~t2f-Ot I��
0.20o00E 0~ 0.1tltE9E-01 1-�
0�t~00ot o1 O.1H90~E-01 I�
0.22000E Ot 0�122712E�Ot , ,
0�2lOOOE Ot O.t00SZ6!'O1 I�
0.2t000E 01 0.12SO40t�02 1� '
0.2lOOOf 0) 0�077~70t-02 1~ . .
0.20000f 01 0. �1 ~0~E-02 ~
0�27000E O1 0.~l1~t1E-01 ~
0.2/OOOE 01 0.309l21E-Oi �
p,29000E ot 0.7027d~E-02 � .
0.lOO0oE 01 0.24790oE�02 ~ �
- 0.lt000E O1 0�20Z9e4E-Ot � �
0.72000E O1 O.tee173f-Oi �
O.SlOOOE 01 0.1s~olZE-02 �
o..+~oooe o, o.+++~+oe-oz � Fig. 4. S3mulatiOn re8ults for factor A .
o.7e000! 0~ o.140e9lE-OS �
O.S~o00E 01 D�e1~l2le-0! ~
O.S~OOOE O1 0.~00l14E'0! � ,
0.71000E 01 0.~0~1aIE-0! �
O.i0000E 0~ 0�S3!l07E'0!
o.v+oooE ot a.2taev7E�oa . � � ,
0.~2000! Ot 0�214N~E-0! �
4.rJ0o0f O1 0�~1~132E-07 ~ '
0.46000E Ot 0.1l07!!E'OS ~ .
O.~f0~0E 01 0.~2J42lE-O! � ~
0.4~oo0E 01 0.1otoJlE-0! . �
O.~1000E 01 0.127lAfE-Or .
0.1lOOOE 01 O.O~f7flE'0~ � ,
o.~roaoE ot o.s~aeoec-o~ ~
o.eooooe o~ o.~l~o�!-o~ ~
- . 64
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
_ FOR OFFICIAL USE ONLY
_ ~he structural and dynami.c capahflities of simulatinn with the DIANA language
are similar to the CSSL language. TIZ.e simplicity and ease of learaing the DIANA
language extend c4nsiderably the group of sp~tem u~ers.
~ Maia characteristics of the laaguage are: ~
simple and clear form of expressions that describe the model;
simple set of statements that allow descriptioa of the majority of dynamic prob-
_ lems represented by differenti.al equations;
execution of a process for a group of changiag parameters;
automatic ~orting of operati,ons;
, result output format Chat allows rapid judgment of the properties of the modeled
system; and
the capab3lity of supplementing the language with new func:ional hlocks in the
form of MACRO, as well as the use of FORTRAN functions and subroutines.
This grogramming language caa he used in various fields of scieace aad technology:
in ~,cieatific research experiments, design, for modeling operation.of designs and
dey3.ces, in iadustrial and economic plannfng to improve methods of organization
~ a~ ~t~oductiun technology. ~
Operating experience confirms the great efficiency of the DIANA language and~
_ system in srolving scientific and technical problems by the metho~ of mathematical
models of tfie real processes, continuous in time.
COP~TGHT: Izdatel'stvo "Finansy i statistika", 1981
- .8545
- . CSO: 81.44I~~138
~ 6
FOR OFFICIAL USE ONLY ~
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED F~R RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
N'UR Uh FICIAL USE ONLY
PROBLEM-ORIENTED LANGUAGE FOR SYSTEI~ OF CaMPUTER
AIDED DESIGN OF MANUFACTUItING
PROCESSES OF MECHANICAL ~CHINING OF PARTS
Moscow VYCHISLITEL'NAYA TEKHNIKA SOTSIALISTICH~SKIKIi STRAN in Russian No 9, 1981
(signed to press 21 May 81) pp GS-73
~Article by V. D.~Tsvetkov, candidate of engineering science (USSR),
A. A. Tolkachev, candidate of engineering science (USSR), and
- I. Cher, engineer (Iiungarian People's Republic)] ~
(Text] Computers are being used ever more extensively in developing automated sys-
tems for teclinological preparatioa of production. Used as source informatioa for
such systems are the part drawing and manufacturing instructions. At the current
level of development of computer technology, the information contained in drawings
caanot be interpreted directly by a computer. Therefore, a basic problem in de-
veloping an automated system is the development of a formalized language to de-
scribe t~e source and intermediate information in a form convenient for input into
a computer, machine processing aad output of design results ia the form of
customary manuf acturing documents.
There are a number of languages and methods of codiag that permit description of
the design and manufacturing charac.terj.stics of a part, such as the material, mass,
dimensions, qualitative and quantitative parameters of the surfaces to be processed,
coordinates of their arrangement in space and others [1-4]. But methods for de-
scribing the function and structure of items and the manufacturiag processes are
still inadequately developed. In contrast to existing systems for coding and the
problem-oriented languages, the suggested language contains advanced facilities for
describing not only the parameters uf the surfaces aad the part as a whole, but
also the shaping, dimensional and precision relationships of the elements of a part
of various levels of complexity.. As algoritbmic ~:velopmental experience shows,
it is precfse~y these relationships that determine the choice of manufacturing
solutions.
The language is based on the methodology of the systems approach to constructing
the objects and processes of desfgn. Any machiae building part is considered a com-
plex part that in accordaace with its structure can be broken down into a specified
number of simpler interrelated elements. .
Surfaces to be machined are~taken as the base structural elements of a part, since
they are directly related to the primary structural elements of Ehe manufacturing
process, the steps.� Tliese.surfaces are subdivided into three types:
. 66
,rOR OFFICIAL USE ONI:Y
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED F~R RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
FOR OFFICIAL USE ONLY
elementary (flat, cylindrical, conic aad others);
type, consisting of frequently encountered combinations of several elementary sur-
faces and to be processed by a type tool (shoulders, slots, grooves and others);
normalized, the shape and dimensions of which are prescribed by correspo~ding
standards (center holes, slotted surfaces a~nd others).
In accordance with this, the first level of forma3.ized description is designed to
describe shapes,dimensions, precision of machining and the physicomechanical
properties of the surfaces listed above.
Description of the Basic 3tructural Elements of a Part (First Level). The maj.ority
of surfaces on machine parts are kinematic, i.e. they can be derived by the motion
of a line (generatrix) in space according to a certain law (directrices). The
shape of the most widespread elementary, type and normalized surfaces will be ~
specified when there are given the axis of the surface i, the directrices m and
the generatrices 1 of the line: ~(i,m,l~).
Various types of surfaces result from the great variety of shapes of generatrices
with a comparatively small number of laws of motion of these lines in space. In
connection with this, the classification glossary of types of basic surfaces is
based on the following features: the law of motion in space of a directrix that ~
defines the class of surface (plane, surface of revolution, helical, etc.); the
~~~e of the directrix that establtshes the subclass (plane, shoulder, slot, etc.)
at~d the type of surface (rectangular, angular, radius, etc. groove); structural ~
- varieties of the surface by length (open, semi-open, closed) (figs. 1, 2). These
_ features determine the design and manufacturing peculiarities of surfaces and
affect to a considerable extent the choice of machining method aad type of cutting
tool.
- The description of the shape of surfaces is a mnemonic designation consisting of
letters and numbers. The surface class and subclass is designated by a short title
in the Russian language, and the type and variety by numbers; for example, a rect-
angular slot w3.th a junction of lateral surfaces on~aradius, semi-open in length is
a PAZ13; a cylindrical surface semi-open in length is a TsILl, etc. Internal sur-
faces are.described with the symbol (for example, a through cylindrical hole
is a-TsILl). This description of the shape of part surf.aces is more graphic,
which accelerates learning the language and facilities use of it. In the process,
the complication of machine processing of iaformation is negligible.
Describing dimensions of type and normalized surfaces has certain features due to
their shape being specified by generalized code ia the language of the first level,
aad not in ths form of a set of elementary surfaces. In connection with this, it
is not passfble to indicate elements tn clear form, the distance between which is
specified by a particular dimension. To uniquely describe dimensions, in the
classification tables there are drawings with dimension conventions. Each parame-
ter of a surface is characterized by its type and numeric value (for example,
DB50, LB40, etc.). ,
In describing basic surfaces, a determiaant is used th at is the set of parameters
uniquely~defiaing the shape dimensioas, finish of machining and deviation from
the proper geometric shape of the surface P and its phy.sicomechani,cal properties.C:
~ N1= (~D,. P, .C}. ' ~
67
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007142/09: CIA-RDP82-40854R040500040033-6
r~x vrr~t,?wL U,~, UNLY
tl~ 1 /loetpsxocnru nsocKUe
_ f ~ 6
~
~ b
~ q ~3 p ~4~ ~ ~5 ~6 ~.7
~e c ~4 i~
c a4 ~
� r~' C~ C.t,
. 1
J
1 2 3 f 2 3 4 S 6 ~ Q / 1 3 4
(9) ;10) ~]~1) ~12 (13 (lY4 (}~5) ~106) .(17) (18) (19 0 ~d~ ;
b Vf
a ~t~ t b � ~o x e ~~ie a. ay~
o ~o ~ a S ! ~e = C
_ j. q a k a o ; 7~ o b p ! 2 a q o a'~ .
p p ; p ~ b y a. ~np 5 b o p Q E o~F
V C~ x~ Q ~ n b d X ~y` C 3~ C N ~ s~y
~o v a. c~ a C a 4 ?n q ~c e c C� ~~�o ~ c� $
~ ~ ~ ~ .
.
~ ~ ~ ~ ~ : ~ : .
_ ` Rr . ~ t. ~ " ' ei e~
e, ~u
~ Oa/ omKaumai 0?era Oa.~ onrKpanra~ Jaxa~~ma~ Ora~yrma ,~rra arla~ai
(23 ~
sva ~
nodapx� ~ f~t ~ ~ pa ~ ~ ~ .
Mocnru l_/ 1,.~ ~G!!l!/
~o a~Nr ? ~ ~
~-~Q ooc~ ~
~
Fig. 1. Diagram of classification and coding for plaae surfaces
Key:
1. plane surfaces 14. trapezoidal
2. plane (pl) 15. one-side trapezoidal
3. slot 16. dovetail
4. guide (p,apr) 17. T-t~haped
5. prism (priz~ 18. wedged
6. pyramid (pirm) 19. regulax polygon
7. ~edge 20. rectangle
8. without shoulder 21. rectaagle with tw~o radius sides
9. with straight shoulder 22. rectangle with a radius side
10. with inclined shoulder 23. type of surface by length
11. rectangular 24. open
12. angular 25. semi-open
13. radius ~ 26. closed
~'OR ~ OFFICIAI, USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2407102/09: CIA-RDP82-00850R000500440033-6
FOR OFFICtAL USE ONLY
2 ~oleazxocmu QpJG(FN11A ~l~l
~2~ � ' -
1 3ne~?exmapxere 2 XaNaexu mapue~at 3 KcHO Ku naoaonar+:re
( xaHm) ( xor~n~
1 2 3 4 S 6 7 e 9 f 2 3 4 '
CS) (6) (7) (9.~ (1 . ~ .1~) ~1~+) (15) ~17) ~
i~ e n ~ s Q
i a~ ~ i i a Q x ~ s ~ ~
~ ~e ~ ne ~$C" �b ~ ,o ' cR c�i�
G v ~
~r ~ a�c tei ~ ~ c~ ~ ~ 6� e ~ c�~ o� $ s = o
? ~
~ ~
O q ` � p o '
a~ ~ NI M! I 4 e`,~,
; s I : _ '
O ~
RI RI
. q
~ ~ Q ' O O
_ ~~1
ei ~
OnRlwnrw ~~Q~ 3a~rPNma,~ od,m OmrD+nrtmt no~yomKaN~QA laxpumo~ . Omrcpumcn OonyomKpormai 3anpNmai
- ~ ~ ~ ~ ~ ~ 4~-
_ ~ ~ ~ ~ - ~ ~
~19 (20) (Zi) (19) (20) (21)'
Fig. 2. Diagram of classificati+on and codiag for surfaces of revolution
Key:
1. surfaces of revolution 12. hyperboloid (giper)
2. elementaxy 13. contoured (fason) '
3. face grooves (kant) 14. ~ectangular
4. longitudinal grooves (kanp) 15. angular
5. face 16. radius .
6. cylinder (tsil) 17. trapezoidal
7. cone 18. one-side trapezoidal
8. sphere 19. open
9. torus (,convex) 20. semi-opeu
10. ellipsoid ~ellips) 21.. closed
11. paraboloid (parab~ 22. tsansient
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
rvn vrr~~,,ttit, u,t. V1VLY
' A developed structure is built by specification of features that ~~escribe the
surface parameters. '
Let us consider the description of parameters of some surfaces on a part shown in
fig. 3:
1= TORETsl [face], DB30/FASKA [chamfer], HB3, FG45G;
2= TsIL2 [cylinderJ, DB30(H6), RA2.5/GALT [fillet], RBS;
3 = TORETs2, DB40; ~
4 = TsIL2, DB40 (F7/GALT, RB4).
!+s
C 9
oj .
Q 30'
. J6 '
' _ .0! 16 .
J~ RS O.f7 R~ ~ 3~4i~ ~
~ pr~f ~ -
r A r ~ / ~
p .~O . . 0 ~ n
~ r S J Q
e p B
/ ~0
/
.~S
0~ c�o 3 - t.s s
~ _ rs 3�? O , sonN.a,~
i ~Q r � '
S `
Fig. 3. Splined shaft (numbers in cir cles are surface numbers)
Descri~tion of the Part at the S~~cond Level. Coasidered at the second level are
complex obj ects (parts or portioas of them) formed by combinati,oa in space of basic
structural elemeats. This level is intended for descriptioa of
the shape of the part; . ~
dimensional relationships between the basic structural elements;
relative position of basic elements (parallelism, alignment aad others~ and the
manufacturing requirements for precision of their relative position ~nonparallelism,
noaperpendicularity, wobble and others); and
gener.al information oa a part or a portion of it at the given level of separatioa.
A part is considered specified in the rectangular system of coordinates, the axes
of c~hich run through the most characteristic elements of the part. For example, for
parts of the "body of revolutioa" type, the system of~coordinates is tied to one of
the ead faces and the longitudinal axis of the part; for plane prismatic parts, the
system of coordinates is situated so that the entire part is placed ia the first
quadrant; for parts ~3.th a complex configuratioa, the coordinate system is tied to
the structural bases of the part.
The "union" and "detach" o~erations are used to describe the shape of a part or
spearate portiaas of it at this level. The union operation for part elements Q1 and
Q2 is deaoted by the symbol Ui and consists in the desiga of the more complex
ohject D, derived by ~oiaing these elements: ~ D= QtU1Qz.
70
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED F~R RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
FOR OFFICIAL USE ONLY
The symbol i denotQS the nature of the interconnection of the elemeats in space:
U1, by the axis (coaxial); U2, by a right angle; U3, by any non-righ~ angle; U4,
tangentially; and U5, by parallel axes of the el~ments.
The detach operatior~ is denoted by the symbol \ i and describes the complete ob~ect
derived as a result of detachiag the element QZ from the volumetric element Q1:
U=Qi\rQs�
Just as in the preceding case, the symbol i denotes the nature of the detach opera-
tion: by the axis (coaxial) is denoted by ~ 1; by a right angle, ~ 2; by an
arbitrary angle, ~ 3; and detaching with parallel axes of elements, \ 4.
The nature of~mutual intersection of part elements Q1 and QZ is specified by the
intersectinn relationship.
The variety of design operations aad relationsbips describe major features of,the
_ part shape that are grasped only when looking at the drawing, without specification
- of which the proper design and manufacturing solution coul.d not be chosen with
co~y�ter-aidea design. _
~$~;uature of the relative position of the basic surfaces in the part coafigu~ation
is specified by the following relationships: coaxiality, el; perpendicularity, e3;
paralTelism, e4; symmetry, es; ~rrangement of elements on a straight line, e~;
arrangement of elemeats in a circle, eg; and membQZShip (incidence), Eg. For exam-
ple, the relationstii~ of parallelism of surfac~s Q1 and Q2 is described in the form
of Q1e4Q2, and symmetry of surfaces'Q1 and Q2 relative to the axis OR by Q1~5(OX)Q2.
The shape of a part as a whole or its individual portions in the language of the
second level is described by the graph ~_(X,~), in which the vertices correspond
to the basic surfaces and the ares to the operations aad relationships that describe
the interrelationship of these surfaces. The grapb caa be represented element by
element in th~ form of relattonship tables. Recorded in the first and second
columns of the table are the incident vertices, aad in.the third column, their inter-
related edges. Table 1 shows the descriptioa of the ahape of a part fragment shown
in fig. 3. Information on the part shape, writtea ia the form of an element re-
lationship table, is convenient for entry iato a computer and subsequent analysis
of it in computer-aided design. ,
Table 1
i-tn e ement -t e ement element
!-A snewear /-q saerexr C ab aro~
relat~'ons"~p
5 g . ~
6 12 ~S
12 ~2
6 7 vt
I 1 10 . nz
FOR OFF[CIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007142/09: CIA-RDP82-40854R040500040033-6
rvn vrria..~p?L U~G V1VLY
The dimensional link bet~een the twp elements Qf,and Qj, which may be points,
lines and surfaces, is specified by the type of dimension u, its numeric magaitude
b a.nd is described by the relationship QiubQ~. .
The letters x, y and z denote linear dimensians parallel to the corresponding
coordinate axes, and the letter 1, those arbitrarily positioned in space. Aagular
dimensions are denoted by the letter and diametric by D or R. Fo~ eacample,
QiX5pQ3 designates a distance of 50 mm along the X axie betweea the elements Qi and
Q~ �
Mathematically, any dimensional chain is described by a graph in which the verttces
correspond to part elements (points, lines and surfaces), and the azcs to the type
and aumeric magnitude of the dimension between these elements. The graph for
dimensional links, just as the graph for the part shape, is represented in the
fo~ of a table of relationships.
Technical requirements for the precision of the relative position of the part sur-
faces are defined by the type of deviatioa ESi and its numeric magnitude 8 and are
descr ibed by the expression QiEHidQ~.
The following designate deviation parameters: EH1, misalignment; EH2, wobble; EH3,
aon-perpendicularity; EH4, non-parallelism; EHS, asymmetry; and EH6, nonplanarity.
For example, a wobble of surfaces Q1 and Q2 relative to the surface Q of no more
than 0.05 is notated by the relationship (Ql, QZ)EH~O.OSQ3. ,
The set of technical requirements for the precision of the relative position, just
as the Iinear dimensional chains, is described by a graph, the vertices of which
correspond to the part elemeats, aad the ares to the type and numeric magnitude of
- the deviation parameters. This graph, 3ust as in the preceding cases, is repre-
sented in the form of a table of relationships. Table 2 gives the description of
the dimensional relationships aad technical requirements for precision of the
relative positioa of the individual surfaces of the part showa ia fig. 3.
Table 2
-t e ementl -t e ement I element relationship
. /�A !/llS/!MT f-11 ~JIlIIlNT
type ~ dimensfon
I 3 X qp
- 5 X 85
~ 9 X .
2 OX EH~ 0.05
The f irst three lines in the table d~escribe a clueter of linear dimensions
specif ied from surface 1, and the fourth, the nonparallelism of surface 2 to axis~
4X.
- FOR OFFICIAL USE ~NLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500040033-6
FOR OFFICIAL USE ONLY
Geaeral informatio~ on the part or its individual elements are characterized by the
set of parameters
~D= (HU, ND, LB, BB, DB~ NB, 1ND~ T0,
. SB, RA, RZ, QD, ~H, DN},
which describe the designatioa HD aud aumber ND of the ~art; its dimensions.LB, BB,
DB and iiB; the material 1~; heat treatment TO'and mechanical properties SB; the
ge~eral treatment f iaish RA and RZ; weight .GD, and magnitudes of chamfers aot
specified ia the drawing ~H; and nonspecified maximum deviations DH. "
The overall formalized description of the part:includes general information on the
part, complete informatioa on the shaping, dimensioaal and precision r.elationships
of the part elements betweea each other arid the parameters of each elementar~,
type and nortnalized surface to be treated. For some~problems in manufacturing
desiga, only part of this information is used; in this case, an abbreviated
formalized description of the parts to be treated is compiled and the separate
levels are integrated far conveaient initial description of the parts. ~
. BIBLIOGRABHY ~
iL. ~Begunov, G. S. et al., "~,pparatus of Geometric Descriptioas for Design Automa-
~tion," in "Vychislitel'naya tekhnika v mashinostroyenii" (Computer Technology
- in Machine Buildiagj, Miask, Izd vo AN BSSR;[Selorussiaa SSR Academy of ~
Sciences], 1966.
2. Goranskiy, G, K., "Sistema kodirovaaiya info~matsi~i pri mashinostrottel'nom
. proyektirovanii" [Information Codiag~System for Machine Building Design],
Minsk, Izd-vo AN BSSR~, No 1, 1965.
3. Nikitenko, V. D., "Podgotovka program~m dlya stankov s pragrammnym upravleniyem"
[Preparation of Programs for Numeric-Controlled Machine Tools], Mosco~a,
Mashinostroyeniye, 1973.
4. Brown, S. A. et al., "Descriptioa of the AET Language," COP4I. OF THE AC,M, No 11,
6, 1963.
COPYRIGHT: Izdatel`stvo "F~inansy i statistika", 1981
8545
CSO: 8144/0138
- ?3
FOR UFFICIAL �USE ~ ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
NUQ UFFICIAL USE ONLY
F.XPERIENCE IN USE OF TII~+ SHARING SYSTEM FOR PROGRAM DEVELOPMENT
Moscow VYCHISLITEL'NAYA TERFIIdIItA SOTSIALISTICHESKIItH STRAN ia Russian No 9, 1981
(signed to press 21 May 81) pp 73-81
[Article by G. Brusdeylins, engiaeer (GDR)]
[Textj Unified System data processing systems are used in two modes: batched and
interactive. Under batched processing, a job is fully described before being aent
for processing. The ~obs thus prepared form a coatinuous stream of entry data t~
� be processed with the aim of optimal loading on the data processing system. Afte;
job entry, the user may not affect the cQUrse of its eacecution and intervene in it,
i.e. does not take part in the procesaing process. The ti.me of processing is de-
, termiaed by the computer center. ~s a result, the user waits a loag time for his
~ ,job and this ia typical for this mode. Wairiag time is calculated from the time
the user submits his job until he receives results. In the iateractive mode, 3obe
are processed in steps by entry of partial ~obs aad output of responses. Coa-
tinual user-system communication is established by using a terminal.
The interactive mode is subdivided into the mode of shared use with access to all
resources and the mode of ahared use with access to specified (limited) resources.
The latter allo~s processing from a terminal of a set of tasks specified by the
system (for example, seat reservations, literature search). The volume of tasks is
~ not restricted by the system so that various subscribers may solve all problems
independently of each other. ~The mode of shared acceas to all installation re-
sources caa be effected by using the componeats of the YeS OS "Time Sharing System"
(SRV) [TSS]. ~
TSS is an extension of the YeS OS operating system. TSS is based on the coatrol
program MVT, aad in fature, SVS will. be the basis.
~ .
TSS and the batched mode are compatible siace in the process the same operating sys-
tem componeats are used. The data formats, notation, access methods and compiling
and service programs match in both modes. Thus, programs intended for batched pro-
cessing whea developed can be rua under TSS without change. Howeti�er, usiag the
BTAM ~bas~ic telecommunications) and GAM (graphic] access methods and their EXCP
, ~execute channel pragram] equivalents is aot possible. A minimum of 512K bytes is
required to place T5S in main storage. Simultaneous operation in TSS and the
batched mode is possible. ~
, 74
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500044433-6
` FOR OFFICIAL USE ONLY
The start of the operation on a termi.nal is called the start of a session. A sub-
scriber communicate~ ~ith the system by using the TSS command language. By using
the command LOGON, the subscriber establishes communication with the system, indi-
cating his identifier and procedure name (the procedure contains a job control
statement). This procedure, supplemented by certain TSS data, describes the TSS
job that is i.acluded in the operation until the end of the session.
Tasks, executed at the subscriber's option, are described by eatry of commands~.
Then the appropriate command processor is loaded iato the TSS area, assigaed to
the subscriber and ~~ecuted. Missing data are requested by the system. The com-
maads are used to check the course of the session and control it, to compile,
execute and debug programs, to manage data and to manage and check the system. If a
sequence of commands is neetled repeatedly, it is advisable to store it in a file as
a procedure that can be called by using the EXEC command. A session is terminated
by the LOGOFF command or by a new LOGON command.
If iaclusioa of TSS was provided for when the operating system was geaerated, the
TSS mode can be started at any time by the start co~and. Af ter this, the operator
can vary the TSS mode or terminate it by using other co~ands. This means that TSS
is simply adapted by subscribers to varying loads. Siace TSS is used for data.
transmission fr~m aad to terminals,by an extended remote access method (TCAM), the
system must contain the message control program [MCPJ that is capable of servicing
ti6th ~SS and other modes of application for TCAM.
After TSS is started, users are granted one or more storage areas of a specified
size for job execution. Rema.ining maia storage can be used to process jobs in the
batched mode. Each TSS area processes several TSS ~obs, though only one subscrib~-
er's job is in this area at any given time. This conditioa stems from the TSS job
staying in this area and being given access to system resources only for a certain
time (a time slice). After expiration of the t~.me slice, the contents associated
~ith the job are copied and stored in direct access storage, and another job is
written to main storage .(OP). This process occurs rapidly; therefore, a subscriber
has the impression that he alone is operating with the data processing [DP] system.
TSS Application Areas. TSS is suitable for use ia many fields. The main five are
considered below.
Application in Science and Technology. Thi~ includes use of the computer bq
engineers, statisticians, mathematicians and other specialists who need not be well
versed in DP. In the process, problems occurring in everyday work are solved and
tElis is done by using a terminal installed in the specialist's working area.
Data Processing. TSS permits creation and change of files, source modules of pro-
grams, data, texts (program documentation, for example) aad procedures.
The capability of changing source modules is especially important in software de-
velopment. Word processing is assuming great importance.
Interactive Programming. Software development from a termiaal in TSS mode is called
iateractive programming. In the process, there is the capability of developing,
modifying and executing programs intended for the batched and TSS modes. The latter
offers eff icient facilities for program debugging. Programmer productivity can be
raised considerably because of interactive programmiag.
75 '
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2407/02/09: CIA-RDP82-00850R000500440033-6
rUK Uh'N'IC;IAL USE ONLY
Job Processing in the Conversati~onal M~ode. Generating jobs under TSS and proces-
sing them in the batched mode is ca].led joh processing in th,e conversational mode.
In the process, conversation is permitted only during job generation and when re-
sults are received. Processing accurs without subscriber intervention. Job pro-
cessing in the conversational mode should be used in compiling programs when
the programs process a large quantity of data aad execution time is very long;
there is no conversation during processing (during translatioa of programs, for
example); and ~
when special technical data media have to be installed.
Interactive Application. Problems, as they are realized with use of a computer
with access to particular resources, are also solved in the TSS mode (for example,
making information available, current accounting of av~ilability, data acquisition).
TSS Application Example. The scientific research center of the Robotron Scientific
Production Combine is developing problem-oriented sof tware for general-purpose use.
The DBS/R data base management system is typical of these software facilities.*
TSS is provided to more efficiently develop such software systems. A TSS test by
a limited number of programmers was:run for a year to gain experience. Res~lts
were used to define agreements for the TSS mode and the hardware composition; the
number of programmers included in the operation was expanded.
A YeS-1040 iastallation with 1024K bytes of main storage was used to try out TSS.
Two hard communications channels were used; a YeS-8564 remote display unit with
nine screens was connected to each channel. At the start, TSS was used six hours a
day. The installation was operated under control of YeS OS/MVT 6.0 and was
available primarily for work on sof tware. The installation also operated inten-
sively in the batched mode. Under these conditions, maia storage was allocated as
follows:
102K system area for resident load modules; message scheduler .
64K message control program (TCAM); TCAM buffer storage
96K TSS control area; TSS buffer storage; resident load module area
130K TSS area; (lOK) local system queue area
482K batched processing area
150K system queue area; nucleus of control programs
Software Developmeat. Prior to the introduction of TSS, all ~obs for software de-
velopment were executed during the day by groups of developers at the computer cen-
ter according to a general wvrk process. At night, this work was done under closed
shop conditions. After TSS was introduced, the following work was accomplished
preferably under TSS:
modification of source modules;
change of data, cataloged procedures and command proceduxes;
debugging of individual programs;
_ translation of small programs; large source modules ~rere translated under batched
processing;
*
Bittner, Yu. and Guenther, M., "DBS/R Data Base Maaagement System for Subsystems
ia the 'Manpo~er' System," in "Vychislitel'naya tekhaika sotsialisticheskikh stran"
(Computer Technology of the Socialist CountriesJ, No 5, M~oscow, Statistika, 1979.
76
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-40850R000500040033-6
FOR OEFICIAL USE ONLY
work oa maintenance of program li.braries; and
~ord ~processing (composition of documentatioa):
The batched m~de includes:
~ overall debugging of software systems;
debugging of programs with a large quantity of data;
traaslation of large programs;
operatioa with magnetic tapes;
compilation of source modules; and
execut~on of debugged programs.
Operation at night contiaues under closed shop conditioas. This method assumes
compatibility of data and results in the TSS and batched modes. Also, the ~,p~zoach
was to use the same program libraries in both modes. TSS supports this
compatibility.
Management of External Storage Units. Availability of high-capacity external
storage units is assumed in the TSS and batched modes to provide access to the
libraries and data needed by many prograu~ers. To reduce this requirement to the
minimum, all required storage was divided into that ~ith continuous and that with
s~-coatinuo.us access. Storage with coatinuous access included the libraries and
~lata pften used by the programmer and which, therefore, had to be available during .
~e sntire computation. Parts of storage referr.ed to infrequently by the program-
mer were allocated only at a certain time; recorded there was information rarely
needed, for example, source modules of debugged programs. All progra~ers use.
libraries that are made available under a coa~oa principle (agreements on names,
estimate of quantities). This principle permits gett~ng away from personal
libraries for subscribers and provides that about fiv.e to ten programmers for a
software system use common libraries in both the TSS and batched modes.
To develop Assembler sof tware, the following sysiem of subdivided files is made
available: library for source modules under dev~lopment; library for load modules;
" proced.ure library which includes command procedures and cataloged procedures for
batched mode; and a liUrary tu hold in-house macro instructions. Use of external
storage is observed from a centralized service console. Subscribers may place
temporary files in maia s.torage.
Program Compilation. Programs are compiled as follows. Statements are written on
forms and punched. Initial translation is performed in the batched mode and the
source module is kept as a compo.nent of a sub~ivided file. If errors occur in
translation, the priatout is used to eliminate them in the TSS mode. The IDIT com-
mana is used to correct errors detected, and the program is *.hen retraaslated.
When the source modules are 1arg~, this can be done by sendiag the job for batched
processing. After its execution, the subscrib~r receives a message (he eagages in
ottaer work ia the meantime) and then checks the di~gnos~ic message file by using
the OUTPUT command. Small source modules are traaslated under TSS. In the pro-
cess, diagnostic messages are output~directly to the terminal. Errors are cor-
rected by using the EDIT command. The printout is used to debug the program.
Af ter the files needed for program executioa are allocated by the AI,LOCATE command,
the progrz~ to be debugged is executed by the ~EST comm~nd. After errors are de-
tected and corrected, thk p,rogram is sent for overa~l testing if it is a component
of a software system, or for final testing in the case of separate programs. Final
77
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
rux urri~iwL u~~ UNLY
tests are performed in the batch~d mode. After successfully passing the test, a
source module is transferred from the development library to the archive library,
and the load module to the production library. Then these programs are deleted
in the development libraries.
Enhancement of Programmer Productivity. Programming is.performed in the following
work stages:
analysis of the problem and draf t of the program (formulation of the job and com-
pilation of the program flow chart);
coding (formulation of the problem solution by using programming language state-
ments);
" program debugging (preparation of instructions in the source language in a form
permitting automatic entry into the computer and storage in files, translation,
assembly and debugging of programs, search for and correction of errors);
documentatioa (compilatioa of application documentatioa);
maintenance of programs (changing and~elimination of errors in debugged programs
(corresponds to the program debugging stage)).
With interactive progra~ing, the highest increase in programmer productivity is
achieved at the stage of program debugging and documentation compilation. This is
- due to the following factors:
reduction in waiting time. The long ~raiting time for job execution typical for e
batched mode is considerably reduced in Lhe interactive mdde because sending the
data to the computer center, waiting for the start of data processing and return of
results are eliminated. Reducing waiting time significantly reduces sof tware
development time;
continuity of work. For a programmer to work eff iciently, it is advisable to have
him continuously engaged ia soft~are development. The long wait for results of
job execution in batched processing forces a programmer to work on several pro-
grams at the same time so as to not waste time. Shifting continuously from one
probZem to another lowers programmer productivity; ~
continuous access to all systiem functions from the work place. The continuous ac-
cess to all system functions from the work place during work time raises a pro-
grammer's enthusiasm for work since he can correct an error in a program right away
_ when it is detected. Continuous access is a prerequisite for continuous work. The
absence of time spent unproductively and the access to all documents needed
improves the programmer's working conditions;
efficient util~.ties. TSS offers a number of efficient utilities, for example the
EDIT command, that are not available ia the batched mode. Dynamic debugging is of
special importance: in thQ batched mode, debugging utilities are static, i.e.
they are programmed in advance, cannot be changed during debuggiag and debuggiag
can be terminated only after a complete run. Dynamic debugging facilities allow
making changes to programs while they are being debugged using information gained
while the program is executing. In the process, runs can be terminated on time
_ or repeated after corre.ction. Experience shows that productivity is raised an
average of 50 percent in the program debugging stage; this matches foreign experi-
ence. This productivity increase in the mo~t laborious stage of software develop-
ment should be used as the basis for more extensive use of TSS for this purpose.
Requirements for Organization of TSS. One reason for tha productivity increase in
interactive programming is the drop in a number of organizational requirements that
programmers are forced to meet in batched processing. For example, the DP
. 78
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/49: CIA-RDP82-00850R040500040033-6
FOR OFFICIAL USE ONLY
installation, which is an auxiliary facility, has to be "subordinate" to the pro-
grammer's working rhythm, and not vice versa. In the ideal, realization of this
requir ement means: each progra~ner has to have his own terminal and receive the
optimal services from the system. Such requirements are unacceptable for economic
congfderations, yet the approach must be to r~duce organizational restrictions
with r espect to the programm.er. This requirement determines the organization of
the TS S mode.
Experience shows that the duration of a debugging session is one to two hours. If
this time is considerably axceeded, work concentration drops off. On the average,
one to two sessions are required per day for a programmer. Consequently, one can
assu~ae that one programmer works no more than three hours per day at a terminal.
During a sessfon, a programmer must not experience psychological pressure from the
need of keeping within the time allocated to him, when a terminal is made available
to a progra.mmer only according to a plan, at a specific time and for a specific
period. This contradicts the requirement to free the programmer to the greatest
extent from "organizational pressure." To meet this requirement, one terminal is
- needed for approximately every three programmers.
A terminal may be installed centrally or decentrally. The centralized in~tallation
in a separate room ensures a high load on the hardware and a suitable environment
_ for con centrated work. Also, in furnishing the room with equipment, health norms
mt~~t be taken into account regarding noise, air conditioning, etc. When the
tieaminal is installed in the work area, less additional space is required, but
the us eful load on the terminal declines.
Adaptation Problems. After system generation, a small group debu~ged it using TSS.
At this stage, the following work was performed:
a study was made of optimal system design to obtain responses quickly;
~ technology was defined ~(agreement on names, time of use, etc.);
standard procedures and commands were developed;
utility programs and working documents for progYammers were compiled.
After this, the programmers engaged in interactive programming went through a short
training course (about six hours). The course covered the general concepts of the
hardwar e, TSS and assignments of th~ plan with a phased breakdown, by which the
. command language could be learned on one's own. Then the first session was held
under the direction of experienced specialists. Al1 necessary operations were
master ed within 14 days.
Supplementary use of interactive programming along with the conventional method for
softwar e development requires the programmer to always be ready to shift from one
problem to another. A skilled programmer considers the capability of focusing on
the main pxoblems as an enhancement of the prestige of his work. Once a staff inem-
ber has become familiar with interacti~ve programming facilities, he works more pro-
dv.ctively, and if these facilities are removed, it is considered a demotion.
TSS us e over a year has shown that it is an effective means for making software de-
velopment more efficient. Therefore, compreheasive expansion of TSS use is planned
= and a"Series-2" Unified System Computer (the YeS-1055) wi11 be used for this. It
is planned to use the YeS-7920 as the terminal.
COPYRIGHT: Izdatel'stvo "Finansy i statistika", 1.981
8545 '
CSO: 8144/0138 79
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500040033-6
l+'UR UFFICIAL USE ONLY
. SZPAK-77 INTEGRATED PROGRAPA4ING SYSTEM FOR COMPLEX AUTOMATION
Moecow VYCHISLITEL'NAYA TEKHNIKA SOTSIALISTICHESKIKH STRAN in Russian No 9, 1981
(signed to press 21 May 81) pp 81-86
[Article by A. Aderek, Master of Science, Poliah People's Republic]~
[Text] Progresa in digital equipment and the evolution in methods of controlling
industrial processes have caused a rapid increase in the number of automated sys-
tems with control camputers (UVM). At the same time, continuous growth in the
cost of software development has been noted. Therefore, digital equipment pro-
ducers and users are very interested in special programming languages that, on the
_ one hand, would lower t:te labor input for programming, and on the other, would be
easily mastered by specialists without programming experience.
T'he SZPAK-77 programming system, developed at the Industrial Insti~ute of Automa-
" tion and Measurements MERA-PIAP, is described in this article.
The SZPAK-77 syatem is intended for the System of S~aall Computers.
The experience of application of the preceding SZPAK system in industry has been
taken into account in the SZPAK-77 syetem.
The SZPAK-77 syatem combines the advantagea of a form language and real-time FOR-
TRAN. By its function, SZPAK-77 is a general-purpose basic sof tware, i.e. can be
used for any set of hardware.
The main purpose of the system is programming control computers that implement
supervisory control in real time in automated process control systems (ASUTP)..
The SZP,AK-77 form programming language in this system is intended for programming
typical algorithms for data acquisition and preprocessing and control. Algorithms
for processing process variables defined by farms can be expanded by using supple-
mentary programa, SUP. The expanded real-time FORTRAN IV F language ~s intended
for preparing them. Supplementary programa communicate with the so-called special
programs, optimization ~rograms for example, that are also written in expanded
FORTRAN. ~ ~
Capabilities of making changes to the system, expanding the system, receiving infor-
mation about variable processPs, etc. are provided by the operator-system inter-
actiev language.
80
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00854R004500040033-6
,
FOR OFFICIAL USE ONLY
Process Vari4bles. The basic unit af the form ~arc?~rammin~t lan~;uaga ia rho pro~�oaA
variable (PP) [PV]. I[ is recorde~ in the system after introduction of the com-
pleted form shown in the figure. The form is modeled on those used in the BICEPS
programming system. This form is f illed out most of ten for programming the record-
- ing.and processing of a parameter measured in ax~ object. In this case, the source
of the process variable is the transducer signal fed to a computer by devices for
communication with an object. The converted value of this signal and its associ-
ated supplementary par ametera such as, for example, restrictions on the maximum
and minimum, are acceasible to user progrems and the operator.
Usually in an ASUTP, the values measured in an object are uaed as the basis for
determining certain integrated technical and economic indicators that describe. the
course of an industrfal process. Forms can be f illed out for these values too,
which then become accessible to the operator by proper names. This is the way
values are handled that are obtained, for example, as a result of performing
chemical analyses in a laboratory, and the operator inputs them manually.
Thus, a proceas variable ia any value measur~d automatically in an object, computed
by a program or input by an operator, for which a S2PAK-71 system form is defined
(filled out).
A p~ocess variable is identified in the system by using a name, which conaists of
four characters in the SZPAK-77 version for the MERA-400 control computer. The
firet character must be alphabetic and the rest numeric. In the following
description, ANNN denotes a process variable name.
Digital Control Loop. PV processing algorithms can be divided into two groups.
The first includes algorithms for monitoring filtration, scaling by physical units
and checking the bounds of the current value of the PV. As a result of their
execution, the curreat PV value is stored in computer storage. The second group
includes processing al,.goriChms programmed in supplementary programs and algorithms
for computing the control process. Let us call the set of algorithms in the
second group and the characteristic parameters for a given PV the digital control
loop (KTsP).
It is especially convenient to single out the digital control loope in the initial
stage of system debugging and operation eince this provides the capability of de-
bugging the soft~rare in stages, debugging in the first place the saf tware for cen-
tral recording and processing of data, i.e. everything except the digital control
loops.
System Variables. Associated with each procesa variable is a number of character-
istic values, for example the current value read out in an object at a given time
and limitations on the maximum and minimum. Let us call these values system
variables (SP) [SV]. They are atored in the form of real numbers, can be read out
and meassred by an operator, and are also accessible to the remaining portion of
the SZPAK-77 software system.
System vaz~iables are identified by a name formed from the name of tihe process vari-
able with which the given SV is associated and two numbers that are the SV number.
System variables are listed ~elow.
81
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2407102/09: CIA-RDP82-00850R000500440033-6
FOR OFFICIAL USE ONLY
G/IqN.K: FEPNCTPA!(HA, O6P~1601KA N YAOABAENHE lTfPEMfNNON RPOyfCCA(l1!!f
' Q~ Ncmovkux ~7n(/.I?pOlA4.YKC, Q-?1,3-nycem, 4-N1C, S-dttfrt~4Nqu0NNGA KQCC[!RQ e~J)
~ 2 BeoJsxaveKaAllIl u~ nyn~ma (!-paap~wex, o_~aRpsw~NJ '
0! Turt d0 ( J-6utma6un~tic~, j.6semcbun~KaA np~paloaryar, 3-cvtmvaKalaA, 0-3pytut)
` ~ ' N4u~lno6aHUt /ll!
o s rw~~r."K4I ! trMV4DlAn
a s Kod zpyrtnat .
IfA4AAb~fb/E QANMe/E ~
o ~ ~ navaarKOt eHavtnae nn
APNCBOfHNf 17YAblA 0/?EPATOPA ~
~ No.~rep nynema ~ntpcmopa, Monep NfC
qN~vOBOf CVH~DIdAXNE
Kod arauoda xeaMmo~eNU., m Kacrtmx, N~ nQK~ma,,~ecmortaaoxeNUe
ANAAOfOBOE C4NTDiBaNHf ~ , Sunur,vucRO3um,aod ~xaK�
~ A'od RBDvOda xtaNmoeaHUa i.=xacetmw,MncKra,a,H~ezo~a, 3vaa~,o~r:s1
~ r,ad o:paHUU~HUU ezodMOt~ curNana
Tuq punsmpauJU(0-6a punempaauw, 1-MfIlO[p0l/MQN l�t0 /IOpAdKG~ ~
li;GEC4E1 NA 2'N3H4fCXNE Ed~1HNybl 1'"~"p�perana,~ Z�to nopABxaJ.4atmoma CpIIa munempi
i r4n paexexuq ntptcvema '
A
~ ~ C 1 nocmoAMne~a rtraeevenra '
, ~
D
0 p
~ i r nq y~umerear.~.~s npu xopptKk~u
, ~ ~ .
~ dAMNb/E J6 060Ao~IXf
~lepuod ~6pa6em,rr~
~ /)porpanna re~vucwtMUA ~MaveHU~ /Ul (tcnu I Ol Janwtaxa 1)
t 4uc~ro Cll~onpedtnatnerz noae~aeom~nan .
IIROBEPKA OPPAHNVENNN '
1 i' MOOMCnONOf .70~OMlNY[ f(Ump00NZ XOX?1CXllIOJ ~1-lRMKh(/11fb/!~ O-p�JOMXM!/N7b/f~ '
' ~ BfD?Xf[ OQAaMIIV(Mtlf JNQV(NqA O~ � ~ ~
Z ~ Mq~MIC OJAQNYV~Np~ JXQYIMIlI /III . . . ' . , . . ~ .
2! KOE CQflQ ~N[7QIlO~~IfhO-6!j Cp~JQ~ 1-Cpf/ KQ OlpCM4VCMkK~
J 0 � Kod ntomfoG lOMN �
1 1 ~JQXCfIMQAIMO dOllslCq74M0/ 4Jfi/NIM(I[ II/IJQ Od4Jl fl/p(fOB OBpQaOlllki[
3 t c.IpMXa C/J!�M QDJI7NIl.T C006WINV(i (f-~AIClfdQp111N1l1 C006f!(fNftR N[ Bydyp7�IIIyOHIQq/~1
YOAA3AE!!Nf qpN BOI,~OQf 3A Ofo.IMNVEHNA � .
j 9 ~ IIR; IfpZMf20 OtpaMWIM(tN (IAO aqft((rmp0I1K KONR17X177N I M/MOproCIpKpH COCp10/NyU,
4 0 ~~JfDt20d 11(I}[[ NtOf7//01! lOMII IIDIM(t0 OlOQM4VfN(fI ~qTO 1iYI0JIpQf((fXfft 4f(mPOIMd ~
4 f ~ nns ~ux�uo o.poN~~.M~,. XOMI~XIIIOI / MOpM7A0~0/ lIOd07K(N(~IJ
~ 2 i~fpllOB a?rt!fl~~`NpTIOit lON11 Nf(XMflO OiJQMVVfMqA
~ G ~II~MOXC{tMOQIMO B017yC/RLMO(0 YlMfMfX4~1 /I(f Jq Ody1/ 1I~Pf(Od 06/JQBOq1Kf(
BbIXOaHb/E QAMNbIf .
J ! I1pOtpaMMC ORptdtAA/Ot((�A yJfpalqlNpt
! 6 Tkrt ae~sod~l0.npupocmolNU, l-no~uuuoMHaru)
s~ lJpedtnexa~ dortycmuno~ k~ntxtxut ynpaeatnuA `
A OEC BbIXOQA
~ ~ Tue eeuoBa (~-al,l-coo6ut~rrue ~n~ ortrpaTOpa,4-otryn~map)
6 1 N+,KCCCtfibl~ Nj RQKI~Ra
PEfYdBfOP ,
~ Adatc petyaArnopa: N=xaeetm~r, ~naK~ma, adptc
6~ ~ A xaea:oleiri Izad~ N~ xaeetmu, A~ naxtma, N+ISOJa
6 uumpo~od i:od:NsxaertmN,Njncx~ma,ntemorrono~wrtxru 6umr
Form for proceas variable processing algorithms in the SZPAK-77 programming
- system
[Key on following page; s~quentia2 line numbers used as keys]
- ~ 82
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500044433-6
FOR OFFICIAL USE ONLY
- F0~l: RECORDING, PROCESSING AND CONTROL OF PROC~SS.VARIABLE (PV)
$V narae
O1 - PV source (1, program; 2, R1; 3, conaole; 4, NIC; 5,, R1 remote cassette)
02 - console input of PV value (1, authorized; 2, not authorized)
03 - PV type (1, bistable; 2, bietable interrupting; 3, meter; 0, others)
04 PV designation
OS PV phyaical unite
06 group code
INITIAL DATA
07 PV initial value
OPERATOR CONSOLE ASSIGNMENT
10 operator console number, NIC number
D LG ITAL READOUT
11 quantization period code, number of cassette, number of pack, bit
location, number of bits sign code
' ANALOG READOUT
1~ quantization period code, casse.tte number, pack number, entry number,
range [V] .
input sign~l restriction code
Z4, filtration type (0, without filtratio~; 1, continuous of lat order;
2, continuoue of 2nd order). Filter frequency cut-off
SCALE BY PHYSICAL UNITS .
15 scale control type
16 A~
17 B
18 ~ scal~ constants
19 D
20 P
21 T PV~s, taken into account dur'ng correction
22 G
DATA ON PROCESSING ~
23 processing period
?4 PV value computation program (if there ia a 1 in O1)
25 number of SV~s defined by user
LIMIT CHECK
26 - normal position of digital contacts �(1, clos.ed; 0, opened)
27 PV value upper limit
28 PV value lower limit
29 - cut-off code (blank, no cut-off; 1, cut-off at limit)
30 - dead zone code
31 maximum permissible change of PV in one processing period
32 - inhibit standard messages (1, standard messages will not be printed)
~
s3
FOR OFFIC[AL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500044433-6
NUK UMN'll;lAL U5~ UNLY �
CONTROL WHEN LIMIT EXCEEDED
39 PV upper limit (then the digital contacts are in an abnormal atate)
40 passage below dead zone of upper limit (then digital contacts are returned
to normal poeition)
41 PV lower limit .
42 passage above dead zone of lower limit
43 PV m~sximum permissible change of PV in one proceasing period
OUTPUT DATA
55 program that defines control
56 - type of output (0, incremental; 1, positional)
57 maximum permisaible control variation ~
OUTPUT ADDRESS ~
60 - type of output (1, R1; 3, meseage for operator; 4, controller)
61 cassette number, pack number
CONTROLLER ~
- 64 controller address: cassette number, pack number, address
65 analog input: cassette number, pack number, input number
66 digital input: cassette number, pack number, bit location
System Variable
Number System Variable Deaignation
00 current value of PV
03 limit on maximum of PV (for PV with two states, the normally
closed poaition of the contacts)
04 limit on minimum of PV
07 maximum permitted variation fn PV value in one processing
period '
14 eupplementary SV's specified by user
. . ,
13 + N N ia the number of supplementary SV~a
Pror,^ss Variable State. Irc a procesa control syatem, a computer acquires data and
isa4es control aignals incident to many industrial apparatuaea that may be faulty
or down for maintenance. Some part of the industrial process may not meet required
conditions. All this has an effect on the atate of proceas variables.
The current value of a process variable may be within limits, thus proper, or beyond.
limits, and thue considered doubtful. The current value of a PV may be read out
automatically or, in the case of a measuring circutt failure, input manually. A PV
may be in a system, but updating of its current value may have ceased, and access
to syatem variables for a user program may be prohibited. .
Indicators with two values have been incorporated in the SZPAK form lar~guage to
allow the capability of sending inf ormation on PV states. PV states ar~ listed
below.
= SL~
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2407/02/09: CIA-RDP82-00850R000500440033-6
FOR OFFICIAL USE ONLY
Process Variable
Designation Value
PV activeness active PV; 1, inactive FV
digital control loop .
activeness 0, active; 1, inactive
PV source 0, PV input automatically; 1, PV input manually
PV correctneas 0, correct PV; l, doubtful PV
digital coatrol loop. ~
~ output mode 0~ with control output; 1, without control output
Programa Supplementing Procesaing of Procesa Variablee. The amount of processing
of proceas variables defined by, forms may be expanded by using supplementary pro-
grema (SUP) that are executed under conditions $$~cified by forms, These programs
are not independent taska in a aystem, but are execured as subroutines of the PV
processing problem. They are identified by a three-character name. If a subroutine
has to be executed under certain conditions, ~ta name must be entered on the
appropriate line of the form. Names of supplemen~ary programa are common for a ays-
tem. Any number of procesa variables can be called to a program.
$~,tipplementary programs are written in expanded FORTRAN. Expanded FORTRAN, on one
h~d, providea the capability of executing operations in real time and controlling
tl~~, Q~sration of a task; on the other hand, it enables acceas to the SZPAK data
bas~ and permits interaction with the remaining parts of the system.
The firat group of subroutines ie determined by the standard language of real-time
FORTRAN. The other subroutines are liated below.
Subroutine Functions Call to Subroutine
read out values of system variables CALL FTTCH (i, n, w, m)
store values of system variables~ CALL STORE (i, n, w, m) ~
read out statuses of a process variable CALL STATUS (i, n, j, m)
store etatuaea of a process variable CALL ASTATUS (i, n, j, m)
process proceas variable out of turn CALL SPECIAL (i, n, m)
output of controls CALL OUTPUT (i, n, w, m)
printout texts (with buffering) C~4I.L PRINT (k, 1, f, e~, e2,
printout texts in IS0-7 code (with buffering) CALL PRETEXT (1, p, t)
Note. Values of parameters: i number of PV or SV
n n~mes of PV or SV
w defines address of REAL type variables where the
~ � valuea read out by the subroutine from the SZPAK
data base are to be moved, vr from where the values
to be stored by the subroutine in the data base
are to be moved
j status worde to.be read or stored from/to the
SZPAK data base
~ m INTEGER type v$riable. After return from the aub-
routine, contains information on methvd of
execution of subroutine, possible errors, etc.
s5
FOR OFb'ICIAL USE ONI:Y
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2407/02/09: CIA-RDP82-00850R000500440033-6
r~uK ur~r~ic:~AL uSE UNLY
k number of parametere
1 nuaaber of printer .
f print format
el, e2, e3 valuea to be printed.
p number of charactera
t text to be printed
Specisl Programs. In the SZPAK-77 eystem, a user can progran? anq algorithm, for
optimization for example, that uses data acquired and proceased by programs pre-
pared in the form tanguage and by supplementary programs. These programs exist in
a system as independent tasks and are called special programs. They ar.e programmed
in FORTRAN with expanaiona identical to the expansione for aupplementary programs.
Operator-Syatem Interactive Language. An operator can communicate with the system
by ueing the coneole for the industrial procees operator or the DZM-180-KSR aerial
matrix printer with keyboard. The operator interacts with the system by inpu.tting
interactive language directivee into the computer according to certain rulea. ~ Each
directive has its own key on the industrial process operator console. If the mes-
sage is aent through the DZM-180-KSR, a three-letter code is used for each directive.
It is asaumed that the ~asic system monitor ie assigned the number 0 and that it ~
authorized for use in reading and changing all data on any PV and starting the sys-
tem. �Other conaoles are assigned sequential nwnbers starting with 1. They can be
used to read all data on any PV, but changes are permitted only for those PV~s
assigned the number of the given console. Operator-system interactive langusge
directives are listed below. `
Code
IDN PV identification ~ VST PV status
PVR PV current value LST digital control loop status
PVH PV upper limit PVS source of current value
- PRL maximum permitted variation of PV OUM control output mode
in one processing period EXT apecial pracessing of PV
OIL maximum permftted vari,ation of REP summary of information on PV
control value ALP PV designation
DAT system variable PAL cancel signaling
OUT digital control loop output value
COPYRIGHT: Izdatel'stvo "Finansy i statistika", 1981
8545
CSO: 8144/0138 ~ ~
'
86
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
FOR OFFICIAL USE ONLY
DOS RT REAL-TIME OPERATING SYSTEM
Moscow VYCHISLITEL'NAYA TEKHNIKA SOTSIALISTICHESKIIC~i STRAN in Ruasian No 9, 1981
(signed to press 21 May 81) pp 86-90
LArticle by A. Shuba, master of engineering, Pbliah People's Republic]
[Text~ In cooperation with the other CEMA countrtea, the DOS RT real-time operating "
eystem has been developed in the Polish People~a Republic for the system of small
~omputers (SM EVM). �
Di~S ~T was designed for the SM-3 and SM-4 minicomputere. As is known, the instruc-
ti+~ ~et for the SM-3 and SM-4 processors providsa direct addressing of 64K bytes
or ~2K words (16-bit) of main atorage.
The SM-4 proceasor permits operation with special main storage control unita
(storage dispatcher) that expand the addresa range to 256K bytes (128K words).
Depending on whether or not there is a need for the storage control unit, it is
possible to generate DOS RT versiona with or wit3~out address tranalation. .
A broad range of peripherals produced ~n the CEMA countries.can be used in the SM-3
and SM-4 systems. These include varioua typea of character terminals and alpha-
numeric printers with or without atorage control; it is poasible to generate ver-
sions for display, induatrial and laboratory interfaces (including the CAMAC inter~-
face, for instance). DOS ItT permits efficient uae of all these devi~es with full
use of their functional hardware capabilities. ~ . ~
DOS RT imposes no major additional restrictions on the field of applicatione com-
pared to the limits involving hardware capabilities. The structure and progrartming
languages supported make the system most effic�ient in operating under the following
conditions and f ields of applicatton: . '
real-time operation. Thia application aseumea acquisition and proceesing of
� measurement data depending on time and ccntrol of.real and simulated physical pro-
cesaes. DOS RT does not rigidly restrict the ncnnber of actiona performed concur-
rently in real time; ~
software development. The operating system permits develppment of programs written
eymbolically, correction of them and tranalation into a load module; it also
provides facilities for program debugging; .
data processing. DOS RT permits operation of d4ta base management programs. Such
programs provide the capability of creating, updating and managing data f iles, such
ae payrolls, inventory reporte, etc. The COBOL compiler facilitates development of
auch programs; .
87
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500040033-6
. r'uK uN~h'~c:iAl. U5E UNLY
numeric calculations. The FORTRAN IV compiler supports the capability of
starti~g and operating programs for engineering and scientific calculatione.
The capability of developing networks has a special effect on the use of minicom-
puters operating under control of DOS RT. Communication interfaces permit con-
necting minicomputers into a network with an equal-rights or hierarchical etructure.
DOS RT may be generated as a system specialized for one or more of the applications
mentioned, i.e. it can function in different modes at the same time. This property
allows adapting a working version of a DOS RT system to a working configuration of
hardware and operating requirementa.
DOS RT has an expanded system for operating with files on disk and tape. The tape
file structure corresponda to the American standard.
DOS RT fully automates creation, retrieval, expansion and deletion of files. A
f ile owner cannot affect a f ile and cannot obtain information on the physical
location of a file on a disk volume.
' The system provides protection of accesa to user files for operatione such as
reading, writing, expanding and deleting. Acceas protection may concern the follc -
ing categories of users: file owner, user group, system programs and user
community.
DOS RT provides the capability of operating with MACRO macroassembler, FORTRAN IV
and FORTRAN IV PLUS, CC30L and BASIC. -
Standard syetem facilities fnclude the MACRO translator and FORTRAN N compiler.
All translators produce an object module. It is common to all languages. This
provides the capability of connecting modules wrttten i:n different source languages
at the atage of task development into one task (a task is a program in a form
suitable for starting).
BASIC is an exception since it has not only the property of compilation, but can
operate in the interpretation mode, i.e. it can directly execute a source program
written in a source file or output from a terminal keyboard.
A programmer interested in using all the propertiea of the SM equipment can use the
MACRO macro assembler.~ ~his Asaembler provides the capability of creating programs
conaisting of processor instructions and of using macro instructions formulated by
the user in a given source module, or in macro system librariea or in a user macro
library. Special translator directives,offer a number of other facilities: ~
conventional translation, sectioning of a program, aegmentation of a program and
others:
- It is we12 known that FORTRAN IV ia used primarily in numeric computations. Imple-
menta~~.ion of FO1tTRAN IV in the DOS RT system also pezmits using tasks written in ~
this language for real time operation. This is possible because of a library of
procedures that organize the time relationship of the task written in FORTRAN and
communication with industrial and laboratory interfaces. The FORTRAN IV compiler
has a number of expansions comp~.red to the atandard language.
88 ~
~ FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED F~R RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
FOR OFFICIAL USE ONLY ,
The DOS RT syetem also includes the FORTRAN IV PLUS optimizing compiler. The
efficiency of the code generated by this compiler is two to threefold greater than
that obtained from the standard compiler. The ahortcoming is its longer compilation
time. It can operate only in a system with address traaslation.
DOS RT has a large set o= application programs (programs f or editing source f iles,
translation and compilation, building tasks, managing libra~ies, makfng changes to
various types of filea, etc.). Tt~ro editing programs are provided in the system:
EOI is an interactive editing program used to create and correct any text file, in-
cluding source programs. Onp can use the standard instruction set and create macro
inatructions f ormed from combining standard instructione. The program permits
editing lines and blocks and supports search and context changes;
SLP is batch editing program used to create and correct and text file by using a
series of instructiona written in advance in a fi1e. Interactive operation is
possible. .
During system operatfon and development of many programs, a considerable portion of
program modules may be used repeatedly in various programs. It is advisable to
store these modules in compact form as library ffles. System programs for building
tasks can directly use modules stored in system and user libraries. Ttao components
at~e provided for this:
LBI~ ie a routine used to create and modify libraries of object modules and Assembler
m~~ro directive libraries; ~
TI~ is a task building routine. It builds a task by using files containing object
modules and system and user object module libraries. Tt~ere are three versions of
this routine that differ by size of main storage needed, operating time and
overlap capabilities.
The system has three components f or correcting and debugging:
PAT is a routine used to make corrections in object modules without having to
translate the entire module;
ZAP is a routine for making corrections to a file in the form of a task without
having to rebuild it. One can also make single cor~ections to each file;
ODT is a routine to check the process of implementation of a debugged task. It is
possible to start the task by parts, modify main storage cells, observe the change
in contents of selected storage cells and inhibit a task.
, The expanded system of files requires facilities offering the user the capability
of managing this system. These f acilities include the PIP routine that enables
performing the following operations with files: creation, duplication, deletion,
expansion, change of name and method of protection, deletion from a catalog and
listing of catalogs. The system also has these file management routines:
FLX. This routine permits transferring files from foreign volumes that have a file
structure different from that of DOS RT;
VFY. Thia routine verifies the correctneas of file structure on a given volume.
One can also make certain corrections to this structure which permits detecting and
often eliminating the consequences of possible hardware errors.
89
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
r~K ~rri~ia~ u~~: UNLY
Other DOS RT capabilities include: batched mode of operation, logging of errors
and testing of hardware. The first of these capabilities is supported by standard
system facilities.
System generation produces a working version of the operating syatem that is
adapted to the specific requfrements and hardware conf iguration. The generatinn
source is the generation data base placed on three cartridge disks with a capacity
of 2.SM bytes (for standard distribution of system generator). This base contains
a certain basic operating system ready for operation that provides the capability
of creating a.working system from the working files existing in the base. Genera-
tion consists of three phases, each of which is an implementation of a certain aet
of operator instructions. This simplif iea the generation process and the operator
need only respond in turn to questions askzd about logic, character and numeric
values.
DOS RT is designed for minicomputer systems. Because of the efficient control pro-
gram, expanded application programs and aystem of files, from a terminal operator's
_ viewpoint, it has many properties of an expanded operating aystem on a large compu-
ter. The system is noted for the ease and efficiency of communication with the
uaer. The system's strongpoint is ~he capability of convenient use of it in real
time when operating with tasks written in both macro assembler and FORTRAN.
COPYRIGHT: Izdatel'stvo "Finansy i statistika", 1981
8545
CSO: 8144/0138
9~
FOR OFFIC[AL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500040033-6
- FOR OFF[CIAL E1SE ONLY
SOFTWARE FOR COI~4NNICATION BETWEEN COMPUTER SYSTEM AND SRECIALIZED PROCESSORS FOR
ARRAY PROCESSING
Moacow VYCHISLITEL'NAYA TEKHNIKA SOTSIALISTICHESKIKH STRAN in Ruesian No 9, 1981
~signed to ~reas 21 May 81) pp 100-104
- [Article by Y. Iva,r,;,va and T. Atanaeov, acientiffc associa~es, People's Republic of
Bulgaria]
[~~~t] Multiprocesaor systems offer'great possibilitiea for increasing the through-
py~ A~ electronic computers. In our opinion, there are two main directions in de-
~yg~,~~#{~g mulitprocessor systems. First is the development of systema of identical
~ p= r~,~ssors linked in an appropriate manner. Th~ speed of such syatems reaches
hundreds of millione of operations per aecond. Howev~r, these systems as a rule
have their maximum efficiency when solving problems su.itable for parallel computa-
' tion.
~ The second is the development of multiprocessor $Xstems from general-purpoae and
specialized processore. This approach, perhape,. does not yield the gain in through-
~ put as in the firat case, t~ut is has a number af aubstantial advantages: it pro-
vides universality of electronic computer syste~p~, offera the capability of in-
creasing throughput as needed, aad permits spec+~?ization of computer systems in
one or more fields of application, etc.
The spectrum of specialized processors developed ~nd under development includes
communication and array procesaors, data base processors, IO proceasors, processora
for processing symbolic infoxmation, etc.
Speci~lized processors for matrix computat~ona,are used to process numeric data
arrays, large in volume, that occur in fields such as signal analysis, matrix alge-
bra, mathematical atatistica, numeric eolving of differential equations, etc. Also,
in a number of applied fields (geologic and meteqrologic research, seismology,
military affaira, etc.), methods are used that are characteristic for the above
mentioned fields.
Specialized processors �or matrix computations.$re usually connected as peri~herals
to a computer syatem, implementing thereby their basic function: freeing the CPU
from the large volume of computations involving data matrices. By using array pro-
cessors, computer system efficiency is raised over ten fold, which depends on the
nature of applied tasks be~ing solved, on the ratiq between the volume of array pro-
cessing and data processing in the CPU, on the speci�ic technical parameters of the
91
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
FOR OFFICIAL USE ONLY
computer system, etc. An array processor can be made accesaible to ueers through
existing programming languages; to do this, software has to be developed to provide
communication between a user program and array processor. However, compilers have
been developed for the majority of programming languages in use, and including
apecial statements to communicate with an array pracessor would lead to making
changes to compilers too. In connection with this, the array proceseor access
method (I~I~) [APAM] or the main control part of it is built as a subroutine and is
activated by a user program by using the standard CALL statement. In the general
case, the CALL statement for matrix operations ras the following paremeters: ~
mnemonic code of matrix operation;
control information (gives instructiona to the APAM on the parameter list structure,
on the method for building and executing the channel program, on processing of
certain interruptiona, etc.);
usually specified for each array (operand or reault) are the starting address in
main storage (OP) or in the array processor atorage, if available; number of ele-
ments in the operation; the step defining the relative position of the elements in
the operation in main storage; the data format defining the representation of num-
bers in the array and certain.requirements for controlling the sign or modification
of the parameters listed above.
When the user requires performance of certain cor.trol operations by the APAM or
operations that monitor the operation of the array processor, the CALL statement
contains parameters differing from those described above.
The main functions of the APAM are:
analysis of CALL parameters and informing the uaer of any err.ors made;
generation of the channel program that implements the request or logically complete
series of requests for matrix operations;
control of the array processor as a aystem resource and dispatching of users
awaiting this resource;
monitoring the execution of the channel program for processing of inetrruptfc~na or
activating user programs used for this purpose.
To perform these functione, the APAM communicates with the user program and the OS
control programs. The APAM raakes use of certain system blocks and tables; there-
fore, i?~ constructing it, standard conevtnions for access to system information
must be taken into account. Defining the status of the APAM as part of the basic
system software depends on a number of factors, for instance on the capabilities of
the specific operating system, the size of the APAM, the specific functions that
the APAM must implement, etc. Described below are two types of array processors;
fundamentally different approaches shaped by the factors Iisted are used in
building methods of accegs to them.
YeS-2335 Array Processor Software. The YeS-2335 array processor is connected to a
system through a channel adapter and is controlled as an ordinary selector channel.
The YeS-2335 has direct access to main storage for fetchfng source data and sfiorfng
matrix operation results. All arithmetic operations in the YeS-2335 are performed
on floating-point data.
The APAM and YeS-2335 provide for execution of the f~lowing oeprations: vector move
and conversion, vector move and conversion of floating-point data to fixed-point,
scalar multiplication, vector element-by-element multiplication, inner product of
9~ ~
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
� FOR OFFICIAL LJSE ONLY
vectors, partial multiplication of a vector, convolutfon of type of multiplication,
sum of vector elements, sum of squares, vector element-by-element addition, squar-
~ ing of vector elements with sign preservation, and the difference equation.
_ The APAM also provides for execution of special-purpose operations:
control operations ("no" operation, branch to channel program);
operatione for synchronization of user and channel programs (building of channel
program, activation of channel program, waiting for end of channel program);
operations for specifying addresses of user subroutines for interruption processing
(determination of s4~routine for processing interruptions during YeS-2335 operation,
determination of subroutine for processing program controlled interruptions);
operation for determining the mode of syntactic analysis performed by the APAM.
The APAM is an Asser~:;ler program. Using conventional asaembly, it can be generated
in different versicns as a function of user requirements and specific operating
system characteristics. Since it is not very big, it is entirely connected to the
user'program written in PL/1, FORTRAN or Asaembler, and functions as an ordinary
subroutine. The YeS-2335 APAM daes not require the user to have a detailed know-
ledge of systems programming. For each user request defined by an appropriate CALL
statement, it builds a channel program and a series of descriptors of arrays-
operands and matrix-result. After execution of the channel program has started.,
the user may obtain control for completing processing in the CPU while the array
processor is executing its tasks concurrently k~ith it.
- The diagram shown in fig. l gives a most general idea of the method by whlch a re-
quest for a matrix operation ia satisf ied. in the YeS-2335.
Software for �Array Processor with Its Own Storage. An array processor with its own
- storage is connected to a computer system as a peripheral through the block-multi-.
- plexer channel. The set of operations performed by it is considerably larger than
~ the YeS-2335 set. Operations are provided for regular and inverse Fourier trans-
forma, square interpolation, Weiner-Levinsox? filter, complex multiplications, etc.
The array prccessor has up to 1M bytea of its own storage. A user can divide it
into a maximum of seven partitions. An independent task can be activated in each
partition. A user has up to 15 registers for each partition. A rather complex
a?gorithm can be implemented to solve a problem by making use of the registers and
- index arithmetic facilities (22 operations).
The APAM for an array processor with its own storage is a large, structurally com-
plex program that is a subsystem of the operating system. It is activated by a
standard CALL statement in the user program written in PL/1, FORTRAN or Assembler,
but only its control part is tied to a user program as a subroutine.
The APAM has,the following main components:
monitor that maintains communication with the console operator all allows execution
of commands from the console to m~nage the resources of the array processor;
control programs that define and execute user tasks in ~he array processor;
EXIT prograln~ that support and control the array processor with its own storage at
the system level.
~ 93
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00854R400504040033-6
, H'UK UH'N7CIAL USE ONLY
(1, 4tHmPanexi~u IIp04[CCOP u OR[PQRI(lIXQA ROMPflfO
~ ,
~~n ~Z ~ UlJEPAIINONNAA CHC7fMA
� , llpotpa~rNa Roa~~oiams.rA
06pa6omKa o6pa6omKU nptpetlaxu~ ~6~ ,
~atps~saxun
IIpOtDaMN� RO.Ib)OfO~IIl~A ~14~
Ayartul crtecKa ~8 ~
napa.vtnrpo~ CALI w~~ Mdy~(......)- - . .
~!NlOQqlIA KQMQI caaayawu~i ontpamop r-~~=t---~ ~ _
1fCXaQeNRA ' i q a
_ ' NOI~ RpOfDQ~MK f[ L rtPOrpa~wva n ~ a
ortvcamtAtri xanpvq --+---r'-__----- s c
i ~-----^-----==r ~ranrDUttt.rv _ j ~ . n a
Bndava coo6wt-
Nuv a6 mrr~r6KVS ~ MOmAUYa Mo.aCnua� MamDuva� I 7C. / w o.
ynpQ~.rtNVt Qana� i ortr~ael onrvaNd pt~yi~aa i q ~ o
. MfMlfti KQNO.fOKO{! L~~ ~ ~ ` 1 3 j�`
nporPa.r.vN (1. ~ 1 � ~ 1 ~ ~
- (5)
~ .
Fig. 1. Functions of the Array Processor Access Method for the~YeS-2335
_ Array Processor
Key : ~
1. CPU and main storage 6. user progrem for interrupt processing
2. operating system ~ 1.~ user program
3. array processor access method 8. CALL name APAM
4. interruption proc~essing next atatement
'S. parameter list analysis, 9. channel prcgram
_ channel program.and array 10. descriptors of arrays
discriptors generation, 11. array-operand ~
issuing ~f error messages, 12. array-reault �
channel program execution control 13. channel adapter
14. YeS-2335 array processor
By using a seriPs of CALL statements, a user program calls to the APAM (with the
_ name XXXX),�and through apecial statements, the user marke the beginning and end of
a task (fig. 2). The APAM defines the task, building a channel program (KP) and
control table (UT) that containa the task coded by a special method. During execu-
tion of the channel program, the control tables and arrays-operands (X and U) are
moved in~to array processor storage. And finally, af ter execution of the algorithm
specif ied by the control tables, the array-result (Y) ie moved into main storage.
The availability of its own storage in the array proCessor and the capability of-
parallel execution of zasks in its partitions provides the.capability of making the
maximum efficient use of a computer syatem when suitable planning of the operation
is performed.
_ 94 ~
FOR OFF[CIAL USE ONLY ~
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2047/02/09: CIA-RDP82-00850R000504040033-6
FOR OFFICIAL USE ONLY
a
~1 ~ l(P+rmpan;r.a:u npoytccop u oneDQmrreNaA rtonAm~
/ ` 5
~ M OD MQM~ 0~?QNdC ~ Coof~dua onfDamopa ,y
_ ~
~ ~ ~DOipa~wa rro.reJOlomtan /i~C
~ ~ _ona~ encNU 1 ~udtxmHar vqcnt~ Ma,yq
CALI x r z x pQ~~pa R~~
CAG( xax~c(.,,.) o~eJ~~MU~, ~ \ ~1~~b~ZOdN1lf
pa3dtna n.! ~ ~ nPotpanxn
CAII x x x x(....) Aatnoiaafaxvt
CAII : zxx ~adava 1 ' otuu6ox
~Atl zxxx~,.,,) ~9~ \ ~ prtp[d[Atx~t~ ' Bcmpo~aA naM~me
~ 7dlAOI `ICmpO� ` ~
CAIi xxxxl....l ~adaval (1~i~rOti BaNAfiY xf x2
CALL xrxx I t/l U1 ~
~
CAIG x x x x(....1 ~ 10 ~ GtrptdlRtx~ct ~ � f y2
i ~adav p
/ %
/ ~ ~.20' 3 �YTf fl1Y
\ 11 yTi
~ / Slnpaintxnt ,l 4 ~1~^' i 13
K/if ~ ~ ~ ~ ernonMixutx~ ~ taBa a ~
xl I~12 ) x2 y1,2 i ~~aJaK - S ~:dlara~~~
/ II '
U
f i U? XnY ~ / ~ ~ 2 1 ~ 6 MATOHVNbuI
Yi \ Y2 j RPOyECCOP C
~~`___~f,~. COQCfAfHH01!
OAM/IbW ~ZL~ ~ '
Ffg. 2. Structure of the ~,rray �Procesaor Accese Method for an Array i'rocessor
with Its OHm Storage
Key:
1. CPU and main storage 14. 'KP2 = channel program 2
2. APAM monitor 15. .OS
3. conunand 16. ~resident part of APAM
4. message 17. exit prograirs
5. operator consol~ 18. r.ecognitfon of errora
6. user program 19. definition of partitions of
7. definition of partition n ~buil~-in storage
8. definition of partition n+l 20.~�~eiinition of tasks
9. task 1 21. control of executioa of tasks
10. task 2 22. 'BMK = block-multiplexer channel
- 11. UT1 = control table 1 23. buil~t-in storage
12. KP1 = channel program 1 24. array processor with its own atorage
13. UT2 = control table 2
COPYRIGHT: Izdatel'stvo "Finansy i statistika", 1981
8545
~ CSO: 8144/0138
_ 9>
FOR OFF[CIAL USE'ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
ranc vrrn.~AL uDr, V1VLY
AUTOMA~'ED SYSTEM FOR MASS PREVENTIVE TREATMENT OF POPULATION (SAI~ff'0)
Moacow i/YCHISLITEL'NAYA TEKHNIKA SOTSIALISTICHESKIiai STRAN in Russian No 9, 1981
_ (signed to preas 21 May 81) pp 127-133 .
[Article by G. Aetardzhiyan, candidate of engineering science, B. Bayev,
scientif ic associate, and B. Raychev, candidate ot engineering science, all from
People's Republic of Bulgaria] ~
~ [Text] Mass preventive treatment of the population is an exc.eptionally importa�
; part of the work of public health serv3ces. It is noted for hiL,h labor-intensi~.-
~ nesa and involvea processing a large amount of varied data. In connection with
this, there is the queation of automating treatment by uaing modern computer tech-
nology. To do this, it is neceasary to aquip departmenta of masa preventive treat-
ment with terml~als, teach medical personnel to work with an automated system,
~ use some of ~he computer resources of a given center, and make available
apecialized sof tware that operates in real time.
; The automated $ystem for amss preventive treatment, SA1~'0, usea an applicationa
' program package (PPP) designed to operate in real time under control of the SUIP
PPP together with YeSTEL and YeS EVM hardware.
The SUIP package, operating in real time, supports a terminal network built on the
- basis o� YeSTEL hardware and a centralized data bsae. SUIP offers facilitiea for
interface (macro inetructions), with the use of wh~i,ch user programa make use o~f,
` the capabilities of this package. SUIP operates in one partition (with an optional
. high priority) cf main ~atorage, ,~nd all application programs operate under ita con-
trol in this same partition., As a result, only part of the computer reaources in
a center are uaed by the syst~m. The other partitions under multiprrgramming ; ~
operation may be used for batched procesaing. ,
The hardware aupported by the SUIP PPP includea:
~ the YeS-8501 electromechanical terminal or tha IZOT 8531-Yel on awitched or
dedicated communication channels; .
the VTS 56100, video terminal or IZOT 7925 togethar with a YeS-7186 or YeS-7187
printer on dedicated channels;
the IZOT 8500 miniterminal on dedicated channels. Thie t~rminal is convenient for
entering numeric data during laboratory tests;
telegraph apparatus on switched or dedicated channel; .
~ a pseudoterminal consieting of the YeS-6012 card reader and the YeS-7033 alpha-
numeric printer.
~ ~ 96
FOR OFFICIAL USE ONLY ~ ~
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00854R000540040033-6
FOR OFFiCIAL USE ONLY
The SAI~O package can operate with the YeS-1020 type computer. Sufficient for it
is the hardw ~re in the standard conf iguration for this computer and the hardware in
the YeSTEL teleprocessing syatem, including the YeS-~401 multiplexer and the r,e-
quired line apparatus to link the remote terminals mentioned above.
, The SAI~'0 PPP consist of one program operat~,ng in t~he batched processing mode
- (autonomously of SUIP) and eight programs operating in the real time.mode under
control of the SUIP telecommunications monitor. The following program functions
are provided in the real *.ime mode:
a) logging and issuing a number for a patient that has ai~rived (registration
function). Registration includes: ~
obtaining a free record from the main file for the new patient to be treated;
entry of last name, first name and patronymic for patient to be treated;
definition of a personal number for the patient to be treated. The personal ~?~inber
is defined by the system and is used to identify the patient;
b) compilation of information on the number and names of people registered in the
system (inf ormatiQn function). Zhree types of queries are possible:
query on the number of registered patients;
query on nemea and SAMPO numbers of registered patients;
query on names and SAMPO numbers of registered patients with specific initials;
c) system-terminal operator interaction d~iring which various pati~nt data are
erttered on disk storage (conversa*ion function). Data are er.tered under cc+ntrol of
a epecial questionnaire prepared in adv~nce by a systems progra^.~~ner. During the
conve~tsation, the o~erator reaponds'to questions, entering a specified set of data
(text and numeric requisites, coded ~slues). When a question on a numeric requisite
- , is asked, the system reporta the limits in which a response ia valid. A special
category of requisites are the so-called coded requisites, each of which may asaume
one or mor~e text values specified in advance. All valid texts are s~ored in a
SAMPO text fil.e that is a svstem theeaur.ns. Each tce~.` is coded with the r~id of a
specified number called the elementary coded value (E&Z) [ECV]. Elementary coded
values may be only nurnbers that meet the following formal r~quirement: only one
- decimal position of the ECV can differ from zero. For example, 0, 1, 'L, 3, 10, 20,
30, 100, I000 may be ECV~s, but 11, 33, 331, 157 and 202 raay not.
Numbers that cannot be ECV~s (i.e. numbers having more than one nonzero decimal
positi.;;:t) are conaidered as c~~mpositions of ECV's. For example, 11 1s consideren a
composition of the ECV's 10 and 1. Each requisite corresponds to one set of ECV's
- that are called proper ECV's. Each proper ECV correaponda to a specific text from
the system thesaurus. A coded requisite can aseume only those values that are pro-~
per ECV's or are compositiona its proper ECV~s. Valid values of coded
requisites are determined again by using validity limits;
d) entry of a value. of a specific requisite determined by a terminal operator
(entry function). When the indicated requisite has already been assigned a value,
the system issues a warning message and displays the value (if there is no protec-
Cion from unauthoriaed reading) befor~e asking a question on the new value. This
nrogram function in many respecte i~ similar to the precQding;
e) compilation of automatic diagnostic evaluation of patient status. This program
' �unction is declared by including tite suitable element in the appropriate question-
_ naire (AI~'Q TYPE-SET). It is us~d to analyze the values of a set a~f d~gital requi-
aites (requisites-arguments) and select one requisite, called the requisite-function.
- 1'he requisite-function must have the functional deaignation "diagnosis SA1~0," be
97
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R004500040033-6
. rvK urri~~n~ u~~: UNLY
; coded and have only two valid values: 1 means the patient has the corresponding
illness, and 0 means he does not.
f) compilation of patient etatua report (report function). Thia program function
is used by the operator to request the system compile a patient etatus report. In
requegting the report, the operator can define what requiaites (with assigned
value, without an assigned value, outside normal limita, with normal values) and
what groups of requisites are to be included in the report. The following groups
of requisites can be named: identification, anamnesis, interpretation, laboratory,
physiological tests, SAMPO diagnosis, physician's diagnosis. The report can be
sent to a specific t~rminal or the name af the external subscriber PRINT can be
used (which is most o�~en a r~iphanumeric printer);
g) deletion of patient data and storage of it in an archive file on magnetic tape.
This progzam function deletes a record in the main file, keeping in the procesa the
accumulated data, entirely for the TAPE function and partially fa~ the TRAC function.
Besides these basic funcr.f.one, the SAM},'0 aystem performa certain apecial functions:
it uses SUIP facilities to limit accesa to certain program functions, providES pro-
tection for reading and writing of individual requisites and queationnaires (used
by the conversation ftiunction), maintaina identif ication of the originator of cer-
tain requisitea (atore~ the initials of the terminal ~perator that entered theee
requisites), and uaea certain internal SUIP mechaniams, for example automatic st~ .
of a function and automat~c output of data accumulated to a given terminal.
Fig. 1 is a simplified diagram of the operating environment of the SAMPO PPP. The
data base of the SA1~0 packag~ determines the structure and composition of data
needed to execute the various program functiona in the system and also their inter-
r~al organization.. The data base includes files (data sets) needed for SAMPO package
functioning and facil.ities to deacribe the mak~up and structure of data in the
files (tables).
SAMPO data f iles are par: of the overall system of files aupported.by the SUIP
package. '!'wo SAI~O files are included in the SUIP~ file management table: the
main file, MAINFIL, and the text file, TEXTFIL.
. The SAI~O systeme programmer has available language facilities (macro instructions)
to generate the eo-calleu 'distributing table that describes the makeup and physical
structure of data in the main file. The distributing table is defined in SUIP as
an applications program stored in the real-time program library. When needed, it is
loaded into main storage by using the SUIP macro inatruction FQWPC TYPE-LOAD.
Each record in the main file (except a"zero record") can be used by the system to
keep data pertaining to a treated patient. Records are formed and released dynami-
cally. The zero record contains information as to which records are being used and,
_ which are free. The main file has the direct organization with unblocked fixed-
length records, which ie selecte~ by the SAMPO systems programmer.
Ordinary records in the main file include .f,ie~lds that are the individual requisites.
98
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R044500040033-6
FOR OFF[CIAL USE ONLY ~
,
E,: ~ ~733 ~ ~4 ) ~ ~
� ~ ~APXNB CAM'10
3~ ~ I
- L..-- __J
~ -
6A3A AANHFI7( CAM~O ~5 ~ TcPMVIHAAbI CAMI10
i MAI N~
I
L ~ C�8501 6~
~ i c Y~13~ ~ ( 7>
~ TE fll ~ . . N30T8531.E1~ ~
~ ~ I ec�esa~ (6 )
~ ~ ~ATC ? N30T83~1,Ef
~ g ~ ,r-;-~
Z~a~r
FCWRPL � ~ ~ ATf C �
~ e~ VTS 58100
~1 ~ .
~~oN6ANdTEKA CYNfI N30T79Y5 EC�7186
. ' ~ ~11 ~ ~ 2 C�7fe;
r
Fig. 1. SAI~O package operating environment
~y : ~
1. SUIP library 8. IZOT 8531.Ye1
2. SAI~'0 data base 9. ATS ~expanaion unknown]
3. YeS-7033 10. A1~GS Eexpaneion unknown~
4. SA1~0 archives 11. ~ IZOT 7925
_ 5. SAI~O terminals 12. YeS-7186 or YeS-~187
6. Ye S-8501 13 . StJ IP
7. TA [telegraph apparatus~ 14. 3AI~0
Fig. 2 shows the format, structure and organi~ation of record N, where N= 1, 2,
, the nwnber of records (of patients ) in the main f ile, and also the requieite
field RX, where X= 1, 2, M, and M is the number of requisites. The record
must be long enough to hold all the requisitee defined in the distributing table
for a main file record.
~ The distributing table (fig. 3) contains information on the initial length, type,
function and other such attribute~ of each field in main file records. This table
ia compiled by uaing a apecial~macro instruction, and ia aseembled and placed in
the real time SUIP library. Each f iled in a mein file recoxd ie described by ueing
an eight-byte element, the atructure of which i.s ehown in fig..3.
The text f ile ia the general system thesaurus. It holds the deaignations of all
requiaites and elementary coded values (ECV~s) and the limits for valid and normal
digital requisites.
99 i
FOR OFFICIAL USE 7NI.fY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
. FOR UF'FICIAL USE Ol~'i.Y
.RECORD N FORMAT
mOP~'A7 3A~MCN N
Pa p~ Px ~ ~n
mOPMAT ~OAf1 PEKBN3NTA Px
~ ~2 ~ 6C P ~ ~ 4AHHbIE PEKBN3NTA (3 ~
NHv1uNA/~b10 erator i~nitials ~
O~EPATOPA p
Fig. 2. Structure of main file record
Ke:Y:
1. requisite field RX format , 3. requisite data
2. BSR [expansion unknown]
IiEADEg BLOCK ~
1 HkiAAbHbIN 6AOK(3A~OnO60K)
0 ~ �
i
' ELEMENT DESCRIBII~ RE~UISITE RX
j 3AEMEHT,O~NCbIBAFOU.~NN PEKB 3NT Px
I
X I I
I
i 0 1 2 3 4 5 6 ~ SANTbI '
~ H ~ ' B~C'TES
F~ �
~'ig. 3. Structure of distributing table
. BY~S
6AHTbl 1 ~J~
KAF04 ~ QAHHbIE ~
KEY DATA
~ i ~ ~
i i
i S 6
6At~Tb~ gyTES2 v
Fig. 4. Structure af logical reco:d in text file
100
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
~ FOR OFFICIAL USE ONLY
All information in the text f ile can be repeatedly duplicated by uaing various
"styles." Here a atyle is poasible mode of operatfon with the SAI~Q system, i.e
operation of the system simultaneously in several languagee, uae of various forms
for asking questions, etc. A atyle identifier hae two bytes.
The text file has the index-sequential organization. Records are blocked and fixed-
length. A logical record has 88 bytes: 7 bytes for the key and 81 bytes for data.
The key takes up the first seven bytes of the logical record. The structure of the
key and data in the text file logical record is shown in ffg. 4.
The text file.may contain records of various atyles. For example, one style may be ~
used for aslcing questions and another for compiling a report. Various atyles can
also be used when the eystem must operate in several languages. The main
(standard) style is style A.
The text file ie created in the batched processing mode (autonomously of SUIP) by
using a apecial program. Z~ro input cards are required for each text f ile record:
a card describing the record key, and a card containing teh data.
The punched card describing the key has special syntax and muat precede the punched
ca~d with the data.
SAMPO includes facilities operating in real time for making corrections to the text
file. Uaed for this are four traneactions that implement the following functiona:
add a new record to the text file, delete a record, replace a record and change a
record. ~
All the SAI~O functiona discussed have been implemented as SUIP transactions. Thus,
the SA1~0 operator uses a relatively elementary and convenient command language to
make requests to the system. Aleo, the SAI~O eyatems programmer has program
f acilizies. Thus, the data base offers great conveniences in implementing all
progr~m functions and expanding the SAMPO systeta.
The data base is noted for ita universality and flexible structure. Formal replace-
ment of the contents of data in the data base (files and dsitributing table) and
keeping the same structure permits using it not only in public health, but also in
other fielus. The sof tware package is delivered in source code; therefore, the
programs can be easily modified.
C.OPYRIGHZ: Izdatel'stvo "Finanay i statistika", 1981
8545 ~ .
CSO: 8144/0138
101
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
H'UK Ul~'N'I(;IAL USE ONI.Y
DATA ACQUISITION AND PROCESSING SYSTEM FOR USSR GAS INDUSTRY
Moecow~VYCIiISLITEL'NAYA TEKHNIKA SOTSIALISTICHESKIKH STRAN in Ruasian No 9, 1981
(signed to preae 21 May 81) pp 133-138
[Article by N. Moletz, engineer, Hungarian People's Republic]
[Text] Cloae cooperation in devel.oping ~oftware for the YeS-1010 computer, especi-
ally in developing and improving application sof tw~re packagea, has been estab-
liehed between the Videoton plant and the Institute for Organization and Computer
Technoiogy in MeCallurgy and Machine Building. A specif ic achievement from this
cooperatian is the.information system for the LSSR Minietry of the Gas Industry
that provides on-line and statietical information on the production, use and
export-import of gas. Technical r~quirementa for the system were defined by
specialists in the USSR Minietry of the Gas Industry (Mingazprom).
Task Description. The system for acquisition and querying of gas information
hgs two functional parts.
The dispatch part provides for acquiaition, storage and processing of on-line and
daily production data, and for presenting entry and calculated data on diaplays and
printers. This information ia used primarily to prepare gas production and dis-
tribution reporta. Data is acquired dail,y and repeatedly (a maximum�of 12 times
a day at this stage) from remote terminals op~rating synchronously with a central
computer.
The most important dispatch data are the preseure at certain points in the pipeline,
temperature and flow rate of tne gas, data on conditions of the compresaor stations,
and periodic production indicators from gas production sites. ~
The commercial part provides data on gas production, distribution and use for rela-
tively long periods (a day, l0~days, a montih, quarter; half -year and a year).
1t?is information makes it poesible to evaluate the operation of the gas industry
and compare plan data to actual. Tkius, the system promotea raiaing develcpment
according to plan of gas production and distribution.
Sys~em Requirementa. In the statement of the problem, the Soviet s'pecialiats de-
scribed in detail the content of the aystem input messages, queries, and principles
of operation. The syatem has ta eupport reception o~~ data arriving from 10 remote
information stations, periodic proceasing of it (every hour), continual capabili'ty
of querying data on~the 12 diaplays or printera connected to these displays, and
continuing management of the daily log of aystem ope'ration.
102
FOR OFFICIAL USE ONLY
i
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2047/02/09: CIA-RDP82-00850R000504040033-6
FOR OFFICIAL Z1SE ONLY
Measages on the display screens are called "standgrd iteras." This means that the
system must display an the screen the data according to previously specified algori-
thma. The display operator doea not have the capability of specifying a new algo-
_ rithm. He can only select from those existing and list, restart or terminate
- data queri~s. ~
Hardware. Hardware requirements were based on two YeS-1010 computers (fig. 1) with
a storage capacity of 64K bytes each and with an identical set of peripherals: one
fixed-head magentic disk atorage unit, two replaceable magentic disk storage unite,
one printer, a paper tape station and a general-purpose interface for data transfer.
In the initial state, 10 synchronous terminals are connected to one computer
through the interface, and 12 asynchronous terminals are connected by modems to the
other one. The capability for expanding the set of terminals had to be prcvided
in developing the sof tware system.
Principlea of Solution. Designing the software system involved solving two groups
of problems. The requirements on operation and reliability are fmportant for system
functioning. LTse of the system oeprating in real time is largely a function of re-
sponse time, i.e. the time that elapses from data query time to angwer time.
S~or~e of a lar~e amount of data ia possible only on packs of disks with mavable
he~ds, the accesz time of which is relatively ].ong. Deaign of data placement
. lat,~ely dependa on the amount of data. When the entire system ia fully used, the
aver~ge amount of data collected per day during~ the 12 polling cyclea is about 2M
bytes. Synchronous lines for data transmission under noiae conditions can operate
at the rate of 600 bits/sec. Theoretically, this requires 4 minutes for each line
in each cycle. However, the work done by an operator on a tierminal also has to be
considered (for exemple, replacement of reels, paper tape, etc.). Therefore, the
full time of one reception is about 10 to 15 minutea.
Because the processing programs can service several files at the same time, their
operation can be considerably speeded up by placing the data to be processed and
that already processed on separate disk units. In this case, the time for search-
ing on diaks and thereby the program run time can be reduced. The expected run
time, considering the data base aizes (def ined ae about SM bytes), is gbout 15 to
20 minutes. During proceasing, the data base contains partially updated data, but
the capability of querying it is provided.
There are two forms of data representation in the syatem. With the first, data does
not need conversion and response tiemis about 1 second. With the second, data can
be put out to the acreeR only after convereion. It is very important that the data
conversion time be short. It is advisable ta uae separate storage units to hold
the data base and the file of responses being generate~. Therefore, the contents
of two disk units are divided. The BASIS pack holds the data filea, and the
~'RANSIT pack holda the input meseages and reaponses formed earlier and during
queries. As a function of the contents aad amount of data, the time for forming
responses is 5 to 15 seconds.
. The second, but no lesa important, group of prob~lems is the development of software
and t;ardware solutions that promote the f2axibility of progremmi~g and reducing pro-
gram debugging time. The problem ia solved on the baeis of a moni~tor for multitask
' lp3
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2047/02/09: CIA-RDP82-00850R000504040033-6
FOR OFFICIAL USE ONLY
VT 340 V ~ 340 VT 340 V I 940
. YTS S6f00 VTS 56f00 ' ' '
COS . COS
~ centra centra
YeS-SOoo YeS ye3 Uer~rr?paneMe~y l(2H~npa~eNbiu eS YeS
- 5052 505Z ~Aouaccop 2 npokeccop f 5052 50SZ eS'S060
processor 2 procesaor 1
Fig. 1. Composition of hardware for the USSR gas induatry information system
proceasing which enables solving aimultaneously up to 96 user tasks. For this it is
sufficient that the displays and synchronous terminals be served by independent pro-
grams, and that complex calculations can be divided into partial tasks and that
final results of the calculationa can be derived by linking these partial taska.
Because of this syateat atructure, for example, there is no problem in controlling
the programs that service the several diaplays and synchronoue terminals, because
this is done by the monitor. By considering certain general instructiona for pro-
grammere, programs can be written and debugged individually, as if the system had
_ no other elements. Flexibility of prograrae is enabled by the macro assembler.
Its value is that def initions common to the various programs (for example, structure
of f iles, distribution of tasks, etc.) are defined in one place; thus, changee
needed for programs already prepared require only retranslation of the source text.
Standard modules (for example error meseages) have been created to facilitate the
work of the programmers.
Taek Solution. Frem the viewpoint of functions performed, the gas induaCry informa-
tion system can be divided ~into three aubayatems: acquieition, processing and
querying of data. The same program is run on both computera. The'composition of
terminals can be identical, thanks to which the nwnber of terminals in use can be
doubled. Fig. 2 ahows a diagrem of the operation of the programs, and the data
stream of one computer.
The data acquisition subsystem tranefers data between the synchronous terminals and
the TRANSIT magnetic disk. Data acquisition may be initiated by a synchronaus ter-
_ minal operator. Message entry is performed by the program serving the synchronous
_ lit~e. This program etores the measage on a minidiak, which in case of erroneous or
damaged measagea does not require deletion of them. At the same time, because of
its rapid operation, meseages arriving simultaneously can be received.~
~ 101~ .
. FOR OFFICIAL USE ONI:Y
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
. FOR OFFICIAL USE ONLY
r
vT$ 56f00 VT$ 56I00 YeS-IOfO ~VT 340 V1 ~40
data tranem~ssion service
6icayxu~rxue neptdaK~ daxxsrz
. ~
- > > 2 3 ~ 8 8
I
.
" ~ i ~ ~ j % ~ i
I ~ ~ ~ ' ~ ~ i
~ ~ ~ I I
~ % 1 1 !0 9
/ % s `\1 ~ ~ .
\ \ ~ ~ lP 6Q
4
s query subsyatem
~ llodcuuneera
i ~ .7anpocoe
data acquisition subeyatem ~
noa~~~mtxa ~6opa
aaxxe,: ~i ty
i ~ _ 12 \
~i~ data processing
subsystem it
!/cdcutme.~ra o6pa6omxu l3
daxxerr
Fig. 2. Structure of information system for r~inis�try of Gas In~uatry
service of synchronous linea 8. ser.vice of asynchronous linea ~
2. entry into YeS-1010 9. cqaanand analyzer
3. YeS-1010 output 10. res~onse generator
4. message analyzer 11. caqmercial procesaing
5. archives 12. programs for printing lista
b. maintenance of fiZea on fixed-head 13. p~inter
disk ~ 14. diep~tch processing
7. maintenance of files on removable data flow
diek [left, TRANSIT pack; right BASIS
pack, data base] logical link between programs
The entered complete message is repeated syntact~.cally by one ta~k common to the
synchronous terminals. In event of an error, the message read together with notea
_ of errors ia returned by the line operator. The latter aenda this meseage back to
the data source. When there is no error, he providea for message s~torage on the
TRANSIT disk; in the process, the line~operator can receive a new message. After
placing the measage for storage, the operator send.s the message to the TF.ANSIT disk
of the other computer through the "computer-compu~ter" task. The message, arrivin~
from one computer to the other, is checked syntactlcal~y; therefore, it can be se~t
for storage right away.
io5
_ FOR OFFICIAL USE QNLY '
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
h'Ult Ur'N't(:IAL USE ONLY
Programs in the processing subayatem are atarted directly by the central computer
operator at the end of the data acquiaition period. While procesaing messages
_ coming in, the progrartis update the data base and prepare the data f iles for query.
The query subsystem is always readq for ~ork; it is activated by dispatchers from
the asynchronous terminals by using commands that call data to the displays. The
command read ia checked syntactically by the system. Any error measage is sent
back to the display. When no errors are detected, the reaponse "Ready" is sent
immediately and af ter the program run, a response is generated and displayed on
the screen. Af ter dieplaying the response, the terminal awaits a new query.
Conclusion. The software system devel~oped meets the parameters specified earlier.
Expansion ia provided for. All capabilities of the monitor (96 tasks) are not yet
exhausted. Af ~er expanding the ayetem with terminals (to 18 synchronous and 24
asynchronous), another 31 programs can be included in the aystem.
For sysCem functioning, it is disadvantageous to have all the hardware being fully
utilized, since if a disk unit goes down, the given system configuration could
operate only by reducing the number of functions performed. Joint operation of
two computers is possible also in this case.
The software system developed has epecial and common elements that are easily .
separated from each other. Af ter replacing special with cQrresponding common
element3, the system car. be used, for example, to service rail ticket booths,
hotel reservations, to monitor production processes. and in other areas where
rapid information service is needed.
COPYRIGHT: Izdatel'stvo "Finansy i statistika", 1981
8545
CSO: 8144/0138
106
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
FOR ~ OFFICIAL USE ONLY
a
ACQUISITION OF DATA ON COMPUTER HARDWARE OPERATION IN GERMAN DEMOCRATIC REPUBLIC
~ Moscow VYCHISLITEL'NAYA TEKHNIKA SC~TSIALISTICHESKtKH STRAN it? Rusaian No 9, 13$1
(signed to press 21 May 81) pp 149-154
[Article by E. Hanisch, engineer, German Democ~$tic Republic~
[Text] In ensuring product quality, a large role is played by the evaluation of the
entire set of data on the products. However, its is easy to analyze product quality
at the development and production stage, buC acquisition and evaluation of data on
~roduCt performance duxitig operation entails cdrisiderable difficulties. This is
_ b~'~su~~ products at the using site are not always serviced ar.d n~onitored~ by the ~
rti~~iufA~cturer. Ccnaequently, ~he user must take part in acquiring date on product
Ferfbrn:ance during operation. The need tc acquire such data is especially impor-
tant in the case of costly and high-Ferformance Cot~puter hardware, since the manu-
- facturer obtains the capability of tyking suita~5le steps and thereby avoiding large
- losses when there is deviation he*_:+een actual and planned relfability indicaEors.
Reaulta of acquisition cf data on groduct perfor~dance during operation ~asically can
~e used in the following areas: .
to define achieved ir.dicatcrs of FrodLCt reliability during operation to ccnfirm
- planned reliability of a given product, to derive reliability indicators being
planned for new and updated producta, to solve general guarantee groblems and to
_ ccnfirm oz reject user claims during the guarantiee geriod;
to study typical malfunctiona to take suitable $tiepg ta enaure and enhance product
reliability and tc i~pYOVe the deaign, technolcgy snd Frodcction of new and
updated FrodLCts;
to plan for the supply of spares; �
to evaluate the efficiency and~qus.lity of asintenance; and
te cc~rFile oFtimal, grecedures for rr.ai.r.tait!ing ~~rcdLCts in gccd hcrking erder.
In cor.necti.en with such bro~d capatilities for Reking use of this data, man.ufactur-
ers aften request a very large emc~r,t cf infcrinatf~n from the user on their products.
As a rule, this leads to makirig excessive demande on the user and as a result, to
failure. However, if the amount of data is limit~d, costs for generalizing the
data by the manufacturer are reduced, and with s~itable organizational measures, .
the information feedback system Gan be adjusted.
An information feedback system used for hardware, eomputers in particular, has been
operating in the German Democratic Republic sinC~ 1~75. Experfence gained in this
time confirms this method is justified in practice. Succeas is due primarily to
107
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
, . r~tc uhr~LiAL. U~~ UNLY
the following reasons:
maximum consideration for user interests, in particular in the design of f orms;
creation of a data collection form that compels the information provider to fill
out the form right at the time the eventoccura;
use df existing organizational forms;
acquisition of local information only in ttie amount needed so as to limit
acqusition costs; ~
providing checks to limit the amount of erroneous data.
Computer hardware is very complex.; therefore, full collection of precise data is
very difficult. It turned out~that it is expedient to gather local data on two
se~a~ate forms, especially with regard to the classif ication of local data pro-
viders (operator, engineer) a~ a computer center. So we use the computer log form
(fig. 1) and the malfunction accounting form (fig. 2) with an attached accompanying
card for an assembly when it is replaced. These forms basically meet the recom-
mendations for use within the Unified System of Computers, but in content they wer~
designed specially for use ir. the GDR. The computer operator fills in the log form.
It is used to gather chro.r.olo~ically data on all operations, on ~all productive and
nonproductive periods and ~~n a11 events that occur. Thus, the data in th~ ~.og not
only def ines the reliability indicators, but is used by the uaer for his own
accounting.
Information on failures (analysis, cause, elimination and location of failure) is
recorded by engineers on a malfunction accounting form. The data on both forms
supplement each other and thus the capability is provided for checking misaing data.
In contrast to the usual collection systems, in our syatem the user compiles a so-
- called monthly report based on the computer Iog. C~ntained in it are information
on the installation configuration and generalized information for determining the
indicators for the reliability of the installation and the individual hardware.
Generalization of the computer log data and compilation of the monthly report have
to be done by the user, because it is not possible to generalize and evaluate cez-
trally the entire amount of computer log data generated by all users at the manu-
f acturing plant. This method also has the advantage that incorrect and missing
data in the computer log can be corrected and supplemented by the user. In the
case of centralized evaluation of local data at the manufacturing plants, correction
of the forms is not possible or the possibility of cc~rrection entai~s large costs.
Feedback is organized as follows. At prescribed scb.edules, the user aends the mal-
function accounting forms and the monthly report to the competent agency for main-
tenance at the manufacturing plant. The maintenance agency checks the collected
forms for completeness and accuracy of data and makes c.orrections when necessary.
Also, an evaluation of the data is made with regard to the specific user conditions;
when necessary, instructions are sent to maintenance teama on steps to take to en-
- sure reliability. The monthly reports,are forwarded to maintenance agen~.ies at a
centralized manufacturing center for further reduction of the data and evaluation
of it. Management decisions are made.according to the results. Then, after suit-
able preparation, information on operating reliability achieved ia sent to manage-
ment staffs, development plants, unit manufacturers, foreign trade agencies and ~
the Coordination Center for the Intergovernmental Commission on Cooperation among
the Socialist Countries in Computer Technology.
108 ~ ~
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
- . FOR OFFICIAL USE ONLY
S ~
~I~Y
~ yktl
E C~ � �
C ~E
a
i~~ ~
. q~ � ~
'8`s~ v
~g W ~ .
~
~ . ~ ~
A ia . m M
Y ti~ ~ ~ ~ ~ .
w
= QYa �
~ ~ ~C ~ d ~
b 7/ ~ 0 M~ . V
~ $ _ i' ~ ~ " ~ ~
� w ~ ~ ~ ~
~ o.~ ~ o0 0 a
~ ~~w~~ a~+�
~!1 b ~ 'd ~ ~ ~ s w
0 ~ d b ~o ~ a d
~ eo ~~d~o ~~b~at,r~r~~ '
i~ o a~ ~ w w++ a, > a~ ao
a~+ a~ a, t~ a, ar a
E d+.+ 6 eo ti++ a? s~ a? ~ oo �
M~ . a o d k.~+ a y a~o a~� ,a
~ p" y O+.+ 'C7 ~ a+ ao d? @ m u e~ O~ ~
~a$ ~
d4~ a . . . . . . . . .
^ g O r-1 N e+1 ~t ~r1 ~p e~ 00
r-I r-I ri r-I rl r-1 r~1 rl ri
skE ~ �
~Y$ �
0�3~ ~ .
. ~
~ ~ .
~
~ w
o ~e
~
~ ~5E
- ~ ~
_
R
3 ~
P'1
~ ~ ~j P, P~1
~ ~ E1 L+ LI O ~p~
= Q. .G A 'Cf F+ .41 ~
� ~ !r ~ ~ ~ ~ ~ G
~O ,.~i ~ ~ ~ U Pi fa a
- ~ .a F+ G e TJ 0 41 C 41
r ~ .rl ~ O 'O
+e E a +~i '~~1 ~ O.^! V CI ~ w O o~~D V
~�k ao$~ua'e a'a ~ ~~~o .
� � "md~a ~"o
e~~ ~ u�~ w w..~ w y d..,
c ~ ~ a ~ d ~ ~
w a~o a c a~ a? a o~ w
d F~ N s ~o a+ d 6~' q~ o a a~
~ uuuuaba~ou~ ~aaa�o
. w
a~
~ . . . . . . . � .
.~1 N cr1 ~7 ~1 ~ ~ I~ a! O~
~ ~ 109
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00854R000540040033-6
H~uK ur~r'ic:tA~ u5~; UNLY
b
0
~ �
~o
Cl C
~ ~
:
E ~ i C ~
V M
' r ~ F y'~ Q
q Y i ~ y ~ ~
i
I ~ y O tl ~ ~ ~
Qn O ~ ~ .
t a~ a c � ~r~l ~ fG
~s C� c� . i~ ~ i~ 4~
~ ~E : C ~ ~ ~
~ ~ a a
a
i e~a � ~ o a E
c�BO ~ C E~ 41 eO ~
i ~ ' ~ 00 C ~ ~ f~r ~
� . ~E;E p ~ ~ .�wa ~ O
�y ~'Y, : .
E a~~~ V C r~W.' r~l ~ 4'i i~.1 .a
k~ ~ ~ w~ ~ ~ O o~~''^ u.
's n .-i tQ w d C ~'L! ~ C~1
sE ~ eQ S O ~ O
~G ~ 'd W YI ~ w !J N rl (p
' O W~: la O 1~ G1 f0 ta ~1
F w o o�-~ w++ a ~ o w o~
: e o a c, ~ o ~ ~a
E 00 4! a.r ~ Cl CJ qM �rl
$ a� $ ~ e~ C C! OD 10 ~ r1 N.G Cl d~ d.Y. C1 V
= E ~ ~ a ~ t~ G~ N bJ C1. 1~ ~A 'C7 ~ C1 V 01
~ $ ~ ~u,�~Hd~~ddoc:~.r+~dua
s 0. ~ . . ~ ~ ~o ~ ro ~ eo a~
E ~ F o
a ~ . . . . . . . . . . . .
~ ~ t t~ N M ~ U'1 ~D ^ 00 C~ O r-I N
a s o 0.~ rl rl r-I r-1 rl r-I r-I r-1 N N tV
_
, ff! p O
, r~r~ i
v
Nr _ C ' v
~r~/ea~~n~ W ~
~ y .~1 ap W
A ~ ~ 3 ~ . .
~ =b . . ~
~
, ~ se � . w .
o ~ N
~ O
~ ~ ~ ~ ~ d E
~ ` rl ,a rl
i ` ~ ~ ' (~4 {.1 LJ ~
~ ~ Y Y n ~ ~ ~ ~ TI
~ ~ d y O a) 00 ~ 4! t~
~ N
a A $ : t ~ ~ � .
J''~ x~ i~ i CL fiE b ~{a la R~~ W
a~i 1~+ 9 @ ~ W d O O r+
3� ~e1 rl ri NQ ld
~ j Y E 7 w TI ~ ~ [ a.7 1.~ i~ C'r ~
~ i~`s a~ r+ N 1~ F+ ~ C ~ C J.r U W
= Y *C ~ IC . rr a d~ IQ ~ d d 7 ~ ~ w Q
E C 00 Q~ 'C C~ 41 00 'O d ~Id Nd ~{d
A C ~r~l ~ ~I v~l C~ C~ ~I d i~ QJ
_ i : ~ L+ ~ E ~ 'L! ~ CJ ~ B O > ~ 8 @ O d ~ ~
e e i O l+ F+ ~ t0 O F+ O E's aJ .~I ~
~ ~ s e$ 1.~ Gl 0~ i~ ~ 01 C~ 'CJ W W'C ~ 4) 'O CL
~ B ~i ~ eI i iJ i.~ a) 4) t~ i~ 41 ~ ~ ~ 1~i 'rd vl
w caa.~ d~ w~v aa~ e~ a? w~ a~ u
~~u ~b ~~o ~ c`~i ~u ab a~ ~vi� ~ a~i ~~b
~ ~
. . . . . . . . . . .
. a4 r-1 N er1 ~t v1 ~O ~ op C~ O r-I
11~ .
FOR OFFICIAL USE ONL'Y ' ~
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
FOR OFFICIAL USE ONLY
Flow o~ Fornns and Data Tasks ,
Ao~.+orc aoxy~vtxnroe u ux~op.wakuu � 3ada~u
. IIoR~io~amt~e SER Fill out forma correctlq
~~~N Compile monthl~r report ~
~wcAVxwa ~ Send oi~t monthly report and malfunction
accounting forins on schedule
Check1for arrival and correctness of forms
w~irc~ m~zxd~recKO,o ,o6eeyxa~aHw? Evaluaf~ with regard � to specif ic"s of
~ ~ 4 � P'cbonraox
individual usera and take steps
.Reduce iaalfunction accounting forin data
by us~ng.~computer ~
UtNmvaswo~aHNrrri nyNxrx t6opa Collect all~ ittforntation
daxxr~s xa Nll�KO,v6r~uaiir�Podfornp~ax Reduce information
Send dt~t resulta
f`i~: 3. Organization of system for monitoring product performance and feedback
information in the Germg~i Democratic 1t~public
_ Keji i .
1. BMZh Lcomputer log form] 4. Robot~on maintenance branch
2. B1JN [malfunction accounting form] 5. data cbllection center at Robotron
3. monthly report NP [sciedtific production] combine
, �
100 ?03
' f!0
�f01 !DO !Of /Oj 1!0 1J6
!00 . ~ 100 10Q'.. "
- !00
~ 7l76 ~ l977 1!7! !!1! , /'o/, . . !!f~ !!77 197d 19f9 1'od
� year qea=
Fig. 4. Increase in YeS-1040 system Fig. 5. Increase in mean time to failure
availability factor of YeS-2640 central proceseor
, Data from the malfunction accounting forms ia ~6iltlenaed by using a computer at the
manufacturer's data center and printed out. Th~ reports are sent to the device
developera and manufact~:rere and to the maintenan~e agericy for failure cause
analysis (fig. 3). .
Tt?e data in the monthly reports and malfunction ac~i~unting forms provide the cap-
ability of making nurt?erous evaluatioris. Let us lis~ juet thos~ applied in the GDR:
monthly account of reliability indicatora achievec3 fot~ individual installations
111
- FOR OFFICIAL USE ~NtY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
. r~uK uH~h'tc;lA1. USE ONLY
(with division by uaers) wiChin the guarantee period to check the indicetors
stipulated in the warranty;
generalization of monthly reports compiled by users for calculating reliability
indicatora for inetallations and individual devices to con.firm a specific reliabil-
ity, ~and the calculation is made from different points of view, for example foir
varioue reporting perioda (month, quarter, half-year, ye~ar, period with a change in
- design and technology, etc.), by different inetallation $ites (export computer,
domestic computer);
-:onaideration of the main direction in the development o~E efforts to improve
qt~dlity;
analysis of the causes of failures of the same type of devices.
The method, uaed in the GDR in recent years, to collect data on the operation of
devices has become a tool to exert a continual effect on improving product quality
and reliability. In contracts made, the user assumes the obligation of taking
part in this effort, which ia a prerequisite for maintaining the warranty.
In the GDR, all Unified System computers ~are subject to continual checking within
the warranty period, and somE.*.imes even af ter it expires. Inforraation is gathered
on the operating performance of all export computere, i~ particular those from the
USSR, in accordance with contract conditiona.
Since the day the system was introduced, much practical data has been gathered~on
product operating performance. As a result, the capability has emerged of con-
cretely determining the.development of efforts on improving product quality and
of evaluating the effectiveness of steps taken to raise quality. The ef~ect of
the feedback system fundamentally depends on the type, quantity and use of
results for solving theoretical problems of reliability and for commercial pur-
poses, etc., and a considerable effec~. is due to the continuous monitoring of the
maintenance of the planned reliability. Comparison of indicatora achieved with
- those planned makes it possible to timely and directly affect the quality and
reliability of installationa where planned indicators have not been reached.
The effectiveness of all steps taken ensures a continual increase in product
quality and stability of reliability. The data on the YeS-1040 installation
availability factor and the YeS-2640 CPU mean time to failure (figs. 4, 5) can be
cited as an example. ~
COPYRIGHT: Izdatel'stvo "Finansy i statistika", 1981 ~
' 8545
CSO: 8144/0138
fi
112
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00854R004500040033-6
~ FDR OFFICIAL USE ONLY
i
i ~
(
~ ~
; .
, ~ .
~
NEW SMALL COI~UTER HARDWARE FROM GERMAN DEMOCRATIC REPUBLIC
I~bscow VYCHISLITEL'NAYA TEKHNIKA SOTSIALISTICI~ESKIKH STRAN in Ruaeian No,9, 1981~
(signed to press 21 May 81) pp 155-161 �
[Article by K. I~eidler, engineer, German Democratic Republlc}
[Text] The SM-1624 microcomputer (Robotron K-1510) is an eight-bit, low-through~
. put computer for handling compu.~ational, testing and control taeks. It features:
deaign and functional modularity and flexibility, i.e. the capability of being
, ua,ed within other equipment, low power coneump~ion, high CPU relial~il~.ty and
~tiall dimensions. '
Because of theae propertiea, the computer is used primarily as a.module in automa-
- tic devicea (measuring, regulating, control and teating), laboratory devices in
test beda, and in teleproceasing devices.
~ The eleanent ba8e in the microprocessor are L9I circuitry for the arithmetic and
logic unit with PMOS technology and additional.TTL el~emente with medium acale in-
tegration. ~
Central procesaor specifications:
width, bits ~8
- instruction execution time, inicrosec. 12.5 46.0 (20.,.000 operatione/sec.)
proceasor cycle, microaec. ~ 4.17
data format, byte 1
, inatruction format, bytes 1, 2, 3
number of basic inetructions 48
size of addreaeable area, .K bytes 16K
addresaing direct logging, direct storage, indirect
storage, immediate
I/0 addresa area 16 addresees for input area, 48 for output
int~errupt syatem one interrupt level, marked; .
� eight prograaimable interrupt eublevels
atorage: main 1K bytes, LSI
ROM ~ ~ 2K bytea, LSI
, The battery module permits retention of data in main atorage when�power ie
, ' interrupted.
. ~ 113 ,
POA OFFICIAL USE ONLY. '
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/42/09: CIA-RDP82-00850R000500044433-6
FOR OFFICIAL USE ONI.Y
The SM-1624 computer common bus ia divided structurally into long and short. Cor~-
nected to the ahort bus are the CPU, supplement to the CPU and devices for inter-
face to the control console, and the PROM programming unit. To the long bus, one
can arbitrarily connect atorage TEZ '_cards] and inter~=ace units that implement
- etandard or special interfacee. ~
One can aleo connect a clock and I/0 units for di~ital dai:o ~or 16 signals.
The SM-1624 microcomputer is a modular unit with a width of 440 cmn for mounting
in racke meeting the recommendations of the IrIICYe [expanaion unknown].
The module ho?ds a maximum of twenty 135 x 170 mm carda. The microcomputer bus
is made in the form of a printed circuit board on a response panel.
The computer sof tware has several components. To prepare programs, there is the
I CROSS as~sembler for the Robotron 4000 family of small computers. Zhere are also
system programs to diagnose errore, programa for program~ning the PROM and
utilities for preparing special programe in epecial operating casea.
~ ~ ~
i
w ~ .
. ;a,,;,..,3.,: ! ` ~ . . ,
. ~
: -~..,.y . ~ ~
. . . , ~ ~
. � '
. - i : ' . , ~ ~ ~
' : ; � ~ +r}~. ~ ~ . . � ' . � � . .
Fig. 1. SM-1626 microcomputer
The SM-1626 microcomputer (Ro~iotron K-1520) is an eight-bit microcomputer (fig. 1)
with a higher scale of integration and higher throughput than ttie SM-1624 model.
The SM-1626 microcmnputer ia used for automating control of production, automation
. of tests, in iaboratory complexea in syateme for proceseing information for
acientific and technical and economic tasks of control and checking.
Carda may be in any place in the central procesaor bus. For emall taeks, one card
with the central proceesor witE~uut a computer bus ie aufficient. The element base
for the microprocessor is LSI circuitry for the a~ithmetic and logic unit with
NMOS technology.
11l~
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
~ - FOR OFFICIAL USE O1~1LY ~ -
CPU specification~:
width, bits 8
inetruction execution time, microaec. 1.6 9.2
system step, na ' 407
data format, bytes 1, 2. . ,
instruction format, bytes 1, 2, 3, 4
number of baeic instructiona 158 ~
size~of addressable area, bytee 64K (can be expanded to 128K bytes)
- I/0 addresa.area 2~56 input addresses; 256 output addresses
fnterrupt system maaked interrupCion; unmasked interruption
control in wait mode available
- level of external.signals meets TTL level
storage: main 1K bytes, LSI, N1~OS technology
ROM ~ ~ 1K bytea, LSI, NMOS technology
counter and time-of -day clock (CTC): '
mode of operation counter and"fime-of-day clock are aet by
program �
number of channels 4
parallel I/0 (PIO):
number of channels 2 bidirectional channele of interface
(4 bita) ~
number of modes of operation 4
_ i~iultiprocessor system: '
~ number of computers joined 1 main and 3 subordinate
maximum data tranafer rate, bytes/s 30K
The structure of the K-1520 microcotnpt~ter ia the maximum configuration that can be
reduced to one central processor without a bue, placed on ~a printed circuit board,
together with an additional maximwn of 4K bytea of stora&e. The control console
is used as an additional means for using the compuLer as an open module during
etart-up for oepration, maintenance and eearching for errora and when testing
programs.
Ttie computer functional unita are placed on 215 x 1.70 nrn carde and can be
asaembled, in the form of an open module. The microproceasar bus is made in the
form of a printed circuit board on the rear panel.
Sof tware for this microcomputer includes a macro.aesembler. To solve uaer prob-
lema, there are a number of etandard modulee, for example I/0 modulea, arittunetic
and converting modules and the atandard mathematical functione. There are com-
pilere for the higher programming languages: CDL, BASIC and COBOL,.
There are different oper.ating systema for the areae of application lieted above.
The SM-6307 (Robotron 1154) alphanumeric printer is a multipurpose device for out-
pnt of data in t~e form of numbere, lettera and special chara~cters ~that operates
with the aingle-digit mosaic printing principle. The~device (fig. 2) ie used mani-
ly in medium-size data proceseing systems as a device for dat~a collection and
� off ice mechanization, a device for output of ineasuring data, and in terminals and
other general-purpose data I/0 unite. .
115
_ FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00854R400504040033-6
rv~c vrri~iwl. u~~, u1vLY
� .
' ~ , ~ .
� i
E,
. r:~ .
n:;
' ~ f. i
V i.
. jY4^
� . f:~.' .
. . � `F!tY.:43 ~ ~ y, y .
~
Fig. 2. SM-6301 alphanumeric printer
Specif ications :
- printing rate, cps 25 in atart-stop mode; 45 in continuous
~ mode
number of characters per line 132
character pattern 7 x 5; expanaion to 10 x 5 is possible
number of characters 64 (96)
space between characters, mm 2.54
s~ace between linea, imn 4.23 . ~
width of print roller, man 385 '
number of prinx copiea 3... 5
paper feed technique roll paper feed (automatic pulling of
.edge~s, manual eetting of form, one linefeed
for roll paper, n linefeed.for perforated
paper)
ink ribhon width 13 mm, capable of black and red
printing
main voltage 220V + lOx
' - 1570
main frequency 50 Hz + 1
power coneumption about 150 VA
The SM-6307 printer contains electronic equipment, power~supply, carri$ge with con-
trol, atep motor, paper feed unit and ink ribbon transport unit. '
The printer can be delivered ae a spearate unit for installation in existing
equipment or as a table-top model with ite own houeing. Dimeneions of the table
model with the paper.feed unit are 744 x 475 x 210 mm.
116 ~
FOR OFFICIAL USE ON~.Y ~
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
' " ~OR OFFICIAL USE ONLY
, .
. ~
~t~::~~ r ' . s' ,-i I ri'
�1,~..; + .S i:s~;d i ~
. �
-.:~o ~ t.' ~ I'I'
. ,^.r ~ - .
s, r ;
~
."S`u,,~ GwF~3~ , .
#%::C~,;Y...
a,:.
,''+�^~i,ps. .
~ l:p'SYdn .
~~j,~~'x 4f K~
~ ~
. _ ~ T~ ~ 1' ` '
ki F~' ~
1
~._i~ , IS ~ ~ � . . �
r , "zy ,~~~-,?~y N
$ ~~y ~5~ 't w~~.~ t y~ti`.~..~s" .
t ` ~ ~ ~T ~ ~S~t_~,Y~ 4
~ ~ ('.''4{M~' y k la ; ^T, -i
y s fi t.:~ 3 .
* Y''~.
, . y., l ~~A ~ .I~i' ~ A,~. ~ .
)L
G ~~t'~ ' ~yy
} ' ~ . L ~N .
~ ' z d~ r~~~~ ~~'~7~~ ~
v s ~ ~~~�r~t ~ ~f4~~ s.5 ~t c t n.~ r.'~,,:i. .
n. ~ '
[~.7:t ....~,~~#`~3~L .
. Fig. 3. SM-6903 mark reader with automatic form feed and eorter
The SM-6903 mark reader (Robotron 1375) is a standalone programmable electronic
- unit for data acquiaition (fig. 3). The device operatea on the~basis of opto-
electronic reading of marks. This method permi~ta automatic~recognition of eource
text characters. When the eource forme are prepared, a marker is used manually to
mark the poaitions that,muet be read and conqerted to particular data automatically.
This device should be applied where a large quantitq og data ie encountered that ia
distinguished by simplicity in coding. This results in savings of resources and
manpower and data is rapidly prepared on media.
Specif ications :
~ traneport rate , 400 mn/e
read rate 2000 4000 forma/hr
etorage and control unit ~ modification of th~ K-1510 micxocomputer
data output YeS-5091 casaette tape recorder
main.voltage 220V + 10~
- 15 ~
power consumption 340 VA
form aize ~ 99 x 148 2i0 x 304.8 nmt
apace between rows. variable, minimum of 5~mn
- ~pace between colwnna ~ 8 mm
marker size about 0.5 x 5 nm
marking medium pencil, black ball point pen, black print
data media magnetic caseette, ECMA [European Computer
Manufacturere Asaociation~ 34, 0.15 inch
117
' FOB OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2047/02/09: CIA-RDP82-00850R000504040033-6
rva~ vrrn.irsa, u~G uPILY
recording method KOI [information interchange code7-TS 97/S
' 11/35 or ECMA 34
recorded code KOI-7 or ECMA 6
a complete checking system ia provided system hardware chacki~,~g (magnetic tsne
- to ensure reliability and che~king of error, read unit, power.interruption);
data forma check (odd marking, ekew of forms,
~ row error);
program facilitiea check (check of ffeld
dependence, check of positions., check of
check numbera, intern~l checks).
The unit is supplied in two ways: hard programmed version (basic unit) and a pro-
grammable veraion (basic unit with expanded storage). Both types can be equipped
with the automatic forms feed and aorter. The mark`reader is a table model and con-
sista of the read unit and tranaport, electrical portian, atorage unit (YeS-5091
casaette tape recorder) and. table. �
Data may also be output on perforated tape using KOI-7 code. The perforator is
positioned aeparately. ~
COPYRIGHT: Izdatel'stvo "Finansy i etatistika~~, 1981
- 8545
, CSO: 8144/0138
118
~ FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
FOR OFFtC~AL USE ONLY
MPL/600 AIrGORITI~IIC LANGUAGE
' Moscow VYCHISLITEL'NAYA TEI~iNIKA SOTSIALISTICHESKIKH STRAN in Ruasian No 9, 1~81
(signed to presa 2I May 81) pp 161-165
[Article by A. Aleksandrov and R. Yefremova, bttCh engineers; and Z1.
Aleksandrova, candidate of engineering science~ gll faom People's Republic of
Bulgaria]
~~e~t] MPL/600 ie a subset of the PL/1 algcirithmic language. The MPL/690 compiler,
~~veloped as a cross-compiler, operates wi.th the Unified Sqstem of ~omputera. The
' ~~t~flNlirn computer conf igu~ation required for the Gompi~ler is : the YeS-2020 CPU,
~~~.l~~i bytes of main storage, the YeS�5052 or Ye5~50&2 dis~c storage unit, the YeS-7031 .
printer, the YeS-6012 card reade~ and the YeS~5012 tape drive.
The compiler consiata of an absblute module cat~loged in the absolute module li-
brary (SYSCLB) and a set of subroutines cataloged iri ttte M-sublibrarri of the source
statement library (SYSSLH). The compiler is w~itten in FORTRAN. Ca.Apiler epeed is
satiefactory: 500 I~L/600 statements are pro~~seed in aSout 20 minu;tea on a YeS-
1020 coznputer .
MPL/600 Feature8 and Capabilitiea. An MPL/600 ~ource progratln ie a series of state-
ments canpiled according to the language synta~tic rules. During compilation, the
compiler ans?;zes the source program etatemenf9 ~.hd translates them into the
Assembler language for the SM-600, called an object program. This program can be
translated by using the Assembler in the 5M-600 ~amily of micra~processors and
executed on a micraproceasor sygtem. The MPL/6t~0 compiler detecta syntactic errors
in the source program and puts out the appropitiat~~meaeagea, after which an equiva-
ler~t~group of SM-600 Asaembler in~tructions is generated for each statement. '
Source program etatemertts muat b+~ punched on car~is in columns 1 to 72. If the
atatement does not fit on one card, it can be eontinued on et~cceeaive cards, except
real Constante. Multiple blanks are ignored by the Eompiler. Blanka in literal
data are preserved. Logic atatements and relational operations muat be enclosed by
two blanks. Each MPL/600 atatement may have a labei. A label has no more than atx
charactera, and the f iret must be a letter. Co~nents may be,included at will in
the program for explanations.
The compiler provides the capability of including in th~ source program groups of
atatements written in Aasembler for the SM-600. ~ach etatement of thia type must
have a$ in column 1. Each MPL/600 atatement perfo~a+e one of the following
119
~
~'OR OF'~'ICYAL i15~ ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00854R400504040033-6
h~uK ur~h~tc;tA~ USE ONLY
- functions: specifies characteriatics'of aource program and type of data with
which the compiler operates, and causes execution of apecified operations. MPL/600
statementa are uaually compiled from key words that are used together with basic
elemente in the language: constants, variables and expresaions. Key worda may not
be used as labels. There are aeveral basic groupa of MPL/600 statements.
Arithmetic statements provide the capability of executing arithmetic operationa
and replacing the current vaiue cf a referenced variable by the result. The fol-
lowing arithmetic operations are permitted in the language:
SHIFT ghift (arithmetic left or right)
IAND conjunctian
IOR disjunction
IEOR modulo 2 addition
X multiplication
/ division
+ addition ~
- aubtraction.
Arithmetic statements are executable atatementa. ~
Control statements give the uaer the capability of controlling the execution of ' ie
object program. The folowing control statemen~s are permitted:
GO TO " unconditional branch atatement
IF conditional branch statement
DO loop statement
~ CALL subroutine exit statement
Control statements are executable.
Declarative statemente are uaed to describe t~he type and of and atorage amount
needed for variablea and arrays. The following data types are permitted:
BIT bit string
BINARY binary inCegera
DECIMAL decimal numbere
SIGNED DECIMAL decimal number with sign
CHARACTER character string.
All variables must be explicitly declared in the source program.
Procedural etatements allow a user to define logically independent program units
that can be compi.led separately or together with the main program.
The MPL/600 language provides the capability of working with three-dimensional
arrays (no more than three subacripts). ~
Special attention is paid to pointers, the use of which facilitates procesaing of
unidimensional arrays to a considerable extent.
The MPL/600 library consista of 32 built-in subroutines, aUou 100 subroutinea for
operating with arrays and over 100 user subroutines.
120 ~
~ FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6
FOR OFFICIAL USE ONLY ~ ~ -
r�
The built-in subro~}tinea are uaed by the compiler in tranalating basic arithmeCic
- operations (addition, subtraction, multiplication, division, decimal arithmetic,
data converaion, etc.). The user can also call these subroutines.
Array operation subroutines are u$ed for basic operationa with arrays: assigtiment;
addition, subtraction, multiplication and divieion of all array elements by a con-
atant; member-by-member addition, subtraction, multiplication aad division of two
arraya with an identical dimenaion, etc.
Subroutines ensure high apeed with the minimal s~ze of required main ator~ge.
Library subroutines are called by using the CALI~ statement and parameters. For
example, CALL ZF50 (P, Q, N), where ZF50 ie thn subroutine name, P is the array of
the result, Q is the source array and N is the dimension of the arrays.
Library aubroutines can be placed in permanent storage. With RAM type main
storage they are not needed. All aubroutines are placed in the source statemenC
library (SYSSLB).in the Unified System.computer, from where they are called as
needed by the compiler, which placea them at th~ end of an object program. They
are compiled in Assembler language for the SM-600.
YeS DOS facilities give.the user the capability of cataloging his MPL/600 or
, SM-600 Assembler programs in the library. Each of them can consist of a practi-
c,~i~.~? unlimited number of instructions. The maximum.number of subroutines in a
~ lib~'ary is 236.
MPL/600 and PL/1 Diff erences. The differencea between these two algorithmic
languages boils down to the difference in notation on the input medium, description
of data, statements and compilation f acilities.
Differences in notation on the input medium: ,
- constants must be written on one line;
a label may have no more than aix characters;
the presence of comments is defined by the characCer ] in front of them;
a$ in column 1 indicates a statement in Assembler for the SM-600.
Differences in data description:
the FLOAT attribute is nqt provided for, i.e. in DEGIMAL and BINARY declarations,
FIXED is implied;
the declaration DECIMAL(m, n) corresponds to PICTURE'(m-n-1)9.(n)9,
lengths of variables must be within the following bounds:
_ Type of Variable m n Unit of Measurement .
BIT 1-8 bit
BINARY 1-2 byte
DECIMAL 1-12 1-12 digit
CHARACTER 1-256 character
where m is the total number of charac tere and n ie the number of digita af ter the
point.
i
- 121
FOR OFF[CIAL USE UNLY
APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500040033-6

APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00854R400504040033-6
rvec ~rri~~wL u~~, u~LY
Differencea in statemente and operatione:
logic operatione and compare operations muat be notated as follows:
NOT negation ~
~ AND logical AND
OR logical OR
EQ =
NE #
. GT
^
GE->;
; LT -