English Version

Rabinovich Zinoviy
???????? ??????????? ????? ?.?.???????? ????????????, icfcst@icfcst.kiev.ua

Pioneer Works of S.A. Lebedev's Kiev Laboratory in the Electo-technical Institute AS USSR and Institute of cybernetics by V.M.Glushkova AS USSR in the Field of Computer Creation and Development.


These developments were carried out mainly in the following directions defined by the types of computers:

1. Universal,

2. Specialized,

3. Managing, wide purpose

4. Intelectualized computers with advanced machine (internal) intelligence

5. Micro computers, wide purpose

6. Intellectualized multimicroprocessor super-computers possesing high and superhigh productivity due to the distributed (multiprogram and parallel) processing of the information.

The first direction was based on S.A.Lebedev's fundamental invention of the electronic digital computer, which first copy was started in research operations in November, 1950 and after being improved was handed over to the Governmental Commission on December, 25th, 1951.

In this report the most significant pioneer developments in the other directions well-known in the USSR and behind its limits are mentioned.

Alongside with the mentioning of historically significant achievements some reasons concerning the further prospects in the development of the computers of the specified above classes are discussed.

Applied developments were supplemented, as a rule, with creation of the theory of the computer which, on one hand, generalized the experience achived while the performans of the computer was held, and on the other this theory opened the further ways and prospects of development of computer science and technics. Theoretical investigations were reflected in the number of monographies, including fundamental ones widely used in the scientific research instituts and high schools.

In general, except for the first works connected with the first and second directions, which were related to scientific
School of S.Lebedev, the others were related to scientific School of V.M. Glushkov, actually to the computer branch of this school. Between participants of both schools there were always friendly business contacts and they shared the same opinion. These contacts did not interrupt even after the death of these great scientists in apogee of their creative activity.

Works in the specified directions are being intensively developed in the Institute of cybernetics by the name of V.M.Glushkov even now, and substantially in the spirit of ideas and the outlines left by V.M.Glushkov. So, for example, in the sixth direction the works on the creation of some computers with high and superhigh-efficiency (super computer) and with clastering architecture obtaining the high level of machine intelligence continue being developed.

Recently the bionical approach to the creation of "thinking" machines on the basis of the developing conceptual notions of the structures and processes of information processing carried out in human thinking is also widely used. Such kinds of notions besides applicational value in an artificial intellect have essential cognitive value in the sphere of researches of the brain, as one of the most important scientific problems of the present.

The further statement of the materiail in the report is carried out within the limits of each specified direction in the historical sequence. And, authors being participants of these rely on well known historical and experimental materials, and on one?s own memories, including already reflected in the literature.

Universal Computers

This direction of works of the
Kiev laboratory of academician Sergey Alekseevich Lebedev is connected with the first page of world?s history of computers?origin and development.

Under the direction of Sergey Alekseevich and with his personal participation in all the stages of work the first computer in the USSR and the continental Europe was created - so-called Small Electronic Calculating Machine ? SECM (????). The actual development and design of the machines began in the laboratory in autumn of 1948. But the most general principles of construction and work of this machine which was subsequently called computer, were invented and formulated personally by Sergey Alekseevich much earlier and there are some certain certificates proving it. Therefore these principles, as well as the architecture of the first uniprocessor computer, called Neiman?s, can be also called Lebedev?s. But all this does not make the role and merits of S.A.Lebedev?s contemporaries smaller as they also contributed much suggesting ideas and principles of construction of the computer and were supervising the first development as well. (In the chronological order I.S.Brook, B.I.Rameev, J.J.Bazilevsky should be mentioned first).

But, anyway the first computer created in the
USSR was M??. In the initial variant it was being developed only for two years - and literally from zero - as, even the bare basic construction at first did not exist. On the 6th of November of 1950 the first start of the machine was already held by the 26-th anniversary of the Great October Socialistic revolution and this fact was mentioned by a special act in Lebedev?s notebook and signed by him personaly. At the given stage of the development of ??? it, as a matter of fact, was executed only in the form of an operating breadboard model, on which the principles of its operation were trained. It helped to improve industrial characteristics and allowed to transform a breadboard model into an effective (to those measures) working computer, quite suitable for regular operation. On December, 25th, 1951 the State commission headed by academic M.V.Keldysh, made and signed the certificate of acceptance of the computer. This date is an officially registrated date of the creation of the first Russian computer and the beginning of its regular operation. Simultaneously with it the further increase of opportunities of the machine, in particular, the introduction of additional memory on a magnetic drum, improvement of input-output, etc. (all these things were already certain innovations) was carried out. Especially important factor was that the architecture of ??? possessed all main basic features and qualitative characteristics necessary for a subsequent creation of highly-effictive computers in a class of universal uniprocessor machines. Limitation of its quantitative characteristics was caused not only by specific conditions of that time, but also by necessity to create the first computer in Soviet Union more quickly.

Thus, the development of
??? was the first basic stage of Lebedev?s activity focused on realization of the main objective of his work, the creation of russian highly -efficient electronic computer facilities. The first computer of this class became BESM which started the well-known series of these machines. Their architectural principles which became classical, in the most initial kind were tested and investigated during the in pre-production operation of computer MESM used thus as an operating official breadboard of the model B??.

It is rather significant, that constructional principles of MESM turned out to be more progressive in comparison with similar ones, carried out in the first English and American computers. In particular, in ?E?? parallel processing of codes in the processor was used for the first time, and in internal language of the machine and system of its management the means of transformation of programs components of were stipulated. ????, actually, became one of the very first computers in the world, adequating to the name a computer.

At once after ???? was accepted by the State commission, i.e. since 1952 it started to be used in regular operation to solve difficult economic and defensive problems, being at that time the only suitable for this purpose machine in the Soviet Union. Among these tasks it is especially important to note the ballistic calculations executed in 1952-1953 by the group of mathematicians under direction of A.A.Lyapunov from Mathematical institute of the Academy of sciences of the USSR.

Alongside with users operation, ?E?? also was widely used for development of methods and programming languages and first information technologies, and also for definition of ways of computer technics development, including the research of an opportunity of the transfering of new computerson to tubeless element-constructive base (B.N.Malinovsky).

Experience of ?E?? creation and work essentially promoted the development of digital computer facilities in the USSR and the greater scientific contribution in this sense was done by S.A.Lebedev's, L.N.Dashevsky?s and E.A.Shkabary's monography . This was published under a heading "confidentially" and was used by experts ??.

The next pioneer and exclusively important work of laboratory in the field of universal computers was the creation of computer "Kiev ". This work began after Lebedeva had finally moved toMoscow. The Laboratory of S.A. Lebedev was at that time headed by academician Boris Vladimirovich Gnedenko. The development of computer ?Kiev> was completed at Victor Mihajlovicha Glushkov's appearancem who became the manager of the laboratory in 1956. The main designer of the development was Naumovich Dashevsky. Computer can be quite considered to be a pioneer work as some new original architectural decisions were applied, such as asynchronous management of the basic devices, group operations with updating the addresses which are carried out above complex structures of data, etc., and also a number of element-constructive improvements, including ferrite operative memory, the external storage on a magnetic drum, input-output of numbers in a decimal notation, etc. (Some of them were applied in earlier created specialized computers- see further).

Computer "Kiev " became the main equipment of Computer center of Academy of sciences in the Ukraine, which was formed on the basis of laboratory. The eqiupment of this laboratory had before consisted of only two machines - ??? and ???. "Kiev ", as well as other machines, was making calculations and conducting scientific researches aimed at the development of the first information technologies, methods of interaction of users with machine, improvement of algorithmic languages of programming, etc. And these works were conducted already during the process of new machines creation and rendered certain influence on their architecture.

In monographies of Dashevskiy L.N., Pogrebinskiy S.B., Shkabar E.A. "Computer Kiev "; Glushkov V.M., Jushchenko E.L. "Computer Kiev ", the specified researches were generalized and were the essential scientific contribution to the development of a computer science.

Computer "Kiev " became famous not only in the Ukraine. Its specially made second copy was used as the basic machine of Computer center of the International institute of nuclear researches in city Dubna.

The specialized computer

Sergey Alekseevich Lebedev was the first to proclaim the idea of creation of specialized electronic calculating machines (????) to solve the systems of the linear algebraic equations iterative methods in which calculations would be completely combined with input of initial data. Thus there was possible an application of the consecutive arithmetic device using special dynamic registers on a magnetic drum on which the basic storage of the machine was carried out. Such a rational major principle of the computer structure, offered by Sergey Alekseevich (a simultaneity of input and calculations and consecutive arithmetics), provided rather small hardware expenses of the computer what was also promoted by the fulfillment of logic elements mainly on feritical transformers and semi-conductor diodes. This circumstance (the quantity of lamps was approximately 10 times less than in ????) was especially underlined in the article published in the American magazine " Datamation " with rather favourable mention about ????. Moreover a collective monography " Specialized electronic ?alculating machine SESM " written by developers of machine Z.L. Rabinovich (main designer ????), JU.V.Blagoveshchensk, R.J.Chernjakom, etc. Was republished in English in the USA. This book was one of the first books about digital computer facilities, translated and published abroad and it supported many actual theoretical-technical positions: smallblock design of ????, which was advanced at that time; absence of usual programming as the necessary modes of calculations were established by batch switches and were carried out by the built in management, initial data and results were expressed in decimal system; modes of the machine as a matter of fact represented various group operations which were carried out not only actually iteration at the decision of systems of the linear algebraic equations, but also, for example, calculation of correlation functions and, in general, performance of various matrics-vector operations.

Thus, ???? already contained in its structure the prototype of the matrics-vector processor which started to be used in the practice of computer construction of a little bit later.

???? was regularly used in Computer center AN USSR, and then it was transferred to some educational institutions for laboratory works of pupils.

Despite the efficiency of operational specialization the dominating development was nevertheless received by universal computers. Thus advantages of such specialization have not been lost, as they were embodied in functional structural units of this computer (for example, matrics-vector processors). But nevertheless specialization of the computer already on the basis of universal machines was not only kept, but also essentially developed, namely in a direction of their orientation to functional applicabilities.

The first semi-conductor digital operating machine of wide purpose

Scientific and technical progress at the end of 50th years of the last century set the task of creation of the computer for management (direct or mediated) of industrial targets, for analysis of processes occuring in the observable environment, etc. The fundamental principle of construction of such computers should have been their system equipment, containing, as an internal software and (that is especially important) structural devices to receive data from objects operated or subjected to the analysis and to conduct direct influence on them or while modelling them in order to study.

The idea of creation of the operating machine of wide purpose was put forward by Victor Mihajlovichem Glushkov in 1957. Creation of such machine - ???? as the first in the USSR was a true pioneer invention which scientific management Victor Mihajlovich has divided with B.N.Malinovsky who became the main designer of the machine.

Wide purpose of ???? was provided by preliminar and accompanying researches of specific features of essentially various operated industrial objects and technological processes which were conducted during its designing. Research of the machine in laboratory and industrial conditions showed its high operational qualities and appropriate efficiency of use that confirmed rationality of a principle of wide purpose for operating machines and allowed to recommend ????, which received the name "Dnepr", for a batch production.

The principle of wide purpose caused following basic characteristics of the operating machine:

- Algorithmically full and, moreover, an essentially developed set of operations in internal language;

- Flexible program management;

- Sectional construction of memory to give an opportunity to choose a desirable memory size;

- An opportunity of connection to the machine of wide purpose of various sources of the entrance information (including analog);

- An opportunity of connection to the machine of wide purpose of printers for a conclusion of received results.

All these structural characteristics were provided in UMSHN "Dnepr". It possessed also rather high for that time constructive and quantitative characteristics and became the first machine in the USSR which received mass use in the process of automation of technological processes, complex physical experiments, objects of military purpose, etc. In Institute of cyberneticsin the Ukraine the hybrid complex first in the country combining operating computer UMSHN " Dnepr " with the analog of the model EMU-10 was developed.

It did not limit the value of creation UMSHN "Dnepr" - as the development of the machine was essentially affected by the development of domestic computing mechanical engineering in general and in particular on the development of cybernetic technics. It was issued within 10 years, than the record of longevity among similar machines has been established. Experience of the creation and usageof UMSHN "Dnepr" was reflected in B.N.Malinovsky's monography "Digital operating machines and automation of manufacture " which was widely used in the designing and educational organizations, and also in the book " There is nothing more dear? "

Intelectualized computers

Necessity and development of mass use of digital computer facilities led to the creation of mini computers intended for personal use in the mode of direct and rather simple interaction of the user with machine, including direct dialogue with it.

Achievement of sufficient universality of such machines, considering essential limitation of their cleanly hardware resources has been provided by flexible microprogram means with which their structures can dynamically adapt to a solved problem.

Under V.M.Glushkov's initiative in Institute of cybernetics the first in Soviet Union mini computer " Promin " with multistage microprogram management was developed. This machine, being started in a batch production, was, as Victor Mihajlovich said " as a matter of fact a new word in a world practice " . Its characteristic advantage was the fuldillment of some often seen procedures, such for example, as performance of macros which microprograms were stored in the storage. As a matter of fact it meant increase of an internal level , i.e. interpreted in process of calculations of language of the machine that relieved users of necessity of programming these procedures. In the second half of the 70th mini computer "????i?? " and "WORLD" (developed in 3 years after computer " ????i?? " - see further) made almost the one third of computes produced in the USSR.

After the pioneer development of UMSHN "Dnepr" and computer " Promin " in Institute of cybernetics in the Ukraine the general line of creation of the computer with the advanced machine intelligence privaled. Some preconditions had already been reflected in these machines. This very term appeared a little bit later, and its introduction eliminated a vagueness of the term "intellectualization of the computer". Works were developed with the promotion of "impudent" (on V.M.Glushkov's expression) ideas of the greater approach of internal (machine) language to mathematical, i.e. giving to it all basic properties of algorithmic languages of programming of a high level. Such internal language assumed structurally supported effective system of interpretation.

This idea was accomplished in the pioneer development of Institute of cybernetics in the Ukraine - series of mini computer " WORLD " and the project of high-efficiency computer " Ukraine ". These works caused rather big resonance and a recognition in the USSR and abroad. Computer " WORLD " was the unique Soviet machine bought in the USA (in order to confirm the Soviet priority in what firm IBM was interested due to some business reasons). This fact is rather characteristic, that the specified development were preceded with the deep theoretical preparation detailing the basic ideas. Singularity of the ideas incorporated in the ?WORLD> also caused far not very unanimous recognition, and on the contrary, sometimes even some fierce disputes.

But as soon as machine " World-1 " (supervisor of studies was V.M.Glushkov, main designer - S.B.pogrebinsky) was created and experience of its operation concerning expediency of the mini computer with microprogram realization of internal language of high level were published all doubts disappeared. However these doubts, in a result of public protection of the project computer " Ukraine ", and also occurrence of similar development in some leading organizations of Russia and Armenia were replaced by confidence of correctness and efficiency of such intellectualization of computer. Its rather high level in the computer family "WORLD ", was provided mainly by all the following architectural characteristics and opportunities of the machine:

- Rather high level of internal language in which machine operations directly corresponded to operators of the user language, and this source language - "analyst" - was specially focused on engineering calculations, including and complex analytical transformations (this was the reason to call this machine

"WORLD " and its specially developed source language - "analyst");

- The developed systems of structural interpretation of internal language with the use of dynamic store automatic devices in memory;

- free accuracy and ranges of calculations by dynamic regulation of quantity of categories of numbers, and also their ranges.

There are some other characteristics of architectural and structural features of the computers " WORLDS " based on hierarchical microprogram management in them which allow to realize the importance of this machine (as well as computer " ????i?? "), as the prototype of the subsequently appeared personal computers. Moreover the WORLDS, under some characteristics of a level of intellectuality essentially surpassed appeared ????.

Recognition and rather high estimation of this significant pioneer development was marked by awarding its authors with the State Premium of the USSR. However, absence during creation of computer " ????i?? " and "WORLD" of appropriate basic constructive and technological opportunities in the USSR didn't allow to transform the WORLDS into structurally perfect personal computers. In the American scientific press it was specified, that modern constructives still did not allow to solve the important problem of increase of a level of internal languages up to algorithmic entrance. And if for machines " WORLDS " it was possible to make, for the highly-efficient computer with parallel data processing it appeared at that time practically impossible. For this reason the project of highly-efficient computer "Ukraine" which though had not been produced in "metal", but was developed in Institute of cybernetics in the Ukraine, could serve as a proof of opportunities and expediency of integration in such computers the internal languages of high level.

The main principles of construction of computer " Ukraine " were the following:

- Semantic identity of its internal language to algorithmic language ?????-60 (as, at that time, to the high level language for universal calculations, and also enriched with logic operations);

- Application of virtual memory dynamically distributed and automatically projected for physical memory;

- Methods of the structural interpretation effectively supported by hardware (including store automatic devices);

- Wide application of methods rapid (including parallel) performance of various operations;

- The prediscretion in regular software of a wide set of complex procedures not programmed by the user, and also subprogramms of realization of typical component of various computing processes.

Also as well as the creation of computer " WORLD " materials for the Project of computer " Ukraine " (chief constructor Z.L.Rabinovich) were dispatched in the form of the published scientific and technical report to the impotrant organizations and it positively affected the development of high-efficiency computer facilities in the USSR. So in the book on computer " Elbrus ", it is directly said, that this super computer had been constructed on two fundamental principles - architecture ????-6 and a principle of structural interpretation of languages of a high level (with the reference thus on V.M.Glushkov, machines " WORLDS " and the Project of machine " Ukraine ").

The world prestige of development "Ukraine" is protected by the copyright certificate with a priority from May, 9th, 1962 (but given out only in 1968 after its detailed discussion and protection according to rather positive expert judgement of academician ?? the USSR A.P.Ershov). It would be important note, that the ideology of development of computer facilities of the USSR did not fall behind the advanced western countries (including the USA), but unfortunately, the technological opportunities were not always available.

In distribution and the subsequent development of the ideas of intellectual development of the computer the publications of some articles and monographies, from which we shall specify two, were very important and valuable: " Computers with the developed systems of interpretation " by V.M.Glushkov, A.A.Barabanov, L.A.Kalinichenko, S.D.Mihnovskiy, Z.L.Rabinovich and " Machine intelligence and new information technologies " by T.A.Gripchenko, A.A.Stognija. The first one was focused on experience of the Project "Ukraine" and in view of its general-theoretical value was widely used in high schools. In the second monography experience of the development of the WORLDS was generalized and methods of an effective usage of the computer in which the high level of machine intelligence plays a significant role.

In the conclusion we shall notice, that with increase of production of the computers with parallel architecture (on the basis of the distributed processing the information) the machine intelligence got new specific features, and that matter was discussed in Z.L.Rabinovich's article " About the concept of machine intelligence and its development " (Cybernetics and the system analysis. 1995,

2, ?.1)

The first domestic family of the micro-computer of wide purpose

By this time there was an opportunity on the basis of achievements of electronic technology to pass to the creation of the micro-computer. For the first time these works were developed in Leningrad?s Designing Buro "Svetlana". The institute of cybernetics in the Ukraine accepted active participation in their creation, namely regarding ideology of construction of its architecture (B.N.Malinovsky, A.V.Palagin). In 1976 in DB "Svetlana" the first domestic family from three micro-computers of wide purpose "Electronics S5" (?5-01, ?5-11, ?5-21) for measurement, the control and management of processes and objects was created. The micro-computer possessed microprogram management for performance of complex group operations (procedures), represented microprocessor complexes with opportunities of their application in various configurations. In the micro-computer the microprocessor with flexible architecture was used. It represented single-crystal 16 digit emulating microprocessor. In the USA this result was repeated later, and our achievement (author A.V.Palagin) in one of the American magazines was called " Russian challenge ".

Considerable value in development of the computer theory was the generalization of results of this development (and also others) in Palagin A.V., Denisenko E.L., Belitskogo R.I.'s monographies, etc. " Microprocessor systems of processing of the information "

Highly and superhighly productive computers with the advanced machine intelligence

For construction and operation of highly and superhighly productive computers the principle of V.M.Glushkov turned out to be rather effective. This principle of macroconveyor data processing was concluded by its definition " that to each separate processor on the new step of calculations such task should be given, which allows it to work for a long time independently without interaction with other processors ". This principle was completely accomplished in pioneer development of multiprocessing system " Macroconveyor " which unlike multiprocessing systems available at that time possessed rather effective property - almost linear dependence of productivity on quantity of processors in essentially big range. System " Macroconveyor " has been created in two production versions of EU-2701 and EU-1766 (productivity of the second with the complete set of processors in use - 256 - was estimated in 128 million ?perations/???). At that time (80th years of the last century) it were probably the most powerful multiprocessing computersin the USSR.

Victor Mihajlovich could not witness these macroconveyor computers ( the main designer was S.B.Pogrebinsky), but he made a huge scientific contribution to its creation in the literature has, which was developed and continued by his followers.

Very characteristic is a circumstance, that aiming at all-round increase of productivity of computing systems macroconveyor systems with rather simple processors, but with their greater quantity, appear to be much more perspective, than a complex on powerful processors, but thus already, with their rather limited quantity. And it was recognized by the founder of such complexes, in relation to which the paradigm of macroconveyor computers is, as a matter of fact, an alternative. And in the further pioneer development of Institute it received a rather essential development on basis of already used perfect microprocessor technics, allowed to create the computer in the form of multimicroprocessor with lots of microprocessors in it. But it required the applications of so-called clastering architecture in which the tasks in calculating process would be assigned on separate clasters, consisting already from no more than several microprocessors.

The first pioneer development in this the was creation of the multimicroprocessor computer of a new class - intellectual solving machine IRM which in itself combined the distributed, automatically operated processing the information with realization of internal language of the high level similar with C++ with entered in it some special graph operations. Actually such combination also defined an accessory of this the COMPUTER to a new class. But keep in mind that the main task of the past and the beginning of a this century which is all-round increase of computer productivity was solved with the pioneer development of the multimicroprocessor super-computer with clastering architecture which was accomplished in Institute of cybernetics of a name of V.M.Glushkov in the Ukraine, a breadboard model of which was computer IRM (supervisor of studies was V.N.Koval, main developer was O.N.Bulavenko).

In this century in Institute of cybernetics of a name of V.M.Glushkov in the Ukraine two models of the specified super-computer - ????-1 and ????-2 with peak productivity accordingly 170 and 350 gigaflops were constructed together with company JUSTAR. In the nearest future the creation of model with productivity up to 4 teraflops is being planned. This way the breakout in the field of computer facilities in Ukraine was accomplished and it allowed to achieve result of a world level. Moreover, regarding machine intelligence its further increase (the supervisor of studies of work is I.V.Sergienko, main designer is V.N.Koval) is being planned.

On the 15-th of March 2006Valery Nikolaevich Koval died. The triumphal success in achievement of his aims demanded an enormous pressure of forces. He literally ?burned down? while working. The main designer of multimicroprocessor clastering computers is now A.A.Jakuba.

The urgency and the further plans for development of these works are noted in available publications:

1. Rabinovich Z.L. " About the concept of machine intelligence and its development ". Cybernetics and the system analysis.-1995, - 2,-?.163-179.

2. Koval V.N., Bulavenko O.N., Rabinovich Z.L.intellectual solving machines as base means of high-efficiency computing systems. ????.-1998, - 6.-?.42-49.

3. Koval V., Bulavenko, Rabinovich Z. Parallel Architectures and Their Development on the Basis of Intelligent Solving Machines // Proc. Int. Conf. on Parallel Computing in Electrical Engineering. - Warsaw (Poland.) - 2002. - P. 21-26.

4. Koval V.N., ?????? V.V.multiprocessor ?????????? systems: planning and realization // the Artificial intellect. - 2004. - 3. - ?.117-126.

5. SCIT - Ukrainian supercomputer project/Koval V.N., Ryabchun S.G., Savyak V.V., Sergienko I.V., Yakuba A.A. // Proc. International Conference KDS. - 2005. - Bulgaria, Varna, 2005.

6. I.V.Sergienko ;V.N.Koval SCIT - Ukrainian supercomputer project // National Academy of Science, Ukraine. - 2005. - 8. - With. 3-13.

The governmental support of the Ukrainian supercomputer project gives the basis to hope, that it will be completely executed and it will be the worthy scientific contribution to keep the memory of V.N.Kovale.