Viktor M. Glushkov | History of Computing in Ukraine
Glushkov

Viktor M. Glushkov

Viktor Mikhailovich Glushkov was one of the founding fathers of information technologies in Ukraine, a genius who was ahead of his time. He is remembered in the history of Soviet IT as the author of digital automation theory, the creator of multi-processor multi-pipeline supercomputers, and the founder of the Institute of Cybernetics of the Academy of Sciences of Ukraine. Thanks to his multi-faceted talent, Viktor Mikhailovich achieved global-impact results in mathematics, cybernetics, computing technologies, and programming, and created his own schools of thought in these areas. He was a brilliant scientist, whose life can hardly be summed up in one book, let alone a single article. During the last 9 days of his life, after waking up from a protracted coma, he summed up his eventful life and career in memoirs, written down by his daughter Olga. These writings allowed us to present a general picture of this genius today.

Viktor Glushkov was born on August 24, 1923 in Rostov-on-Don, in the family of a mining engineer. Five years later, his family moved to the mine named after Artem, near Shakhta town. When he was in the fifth grade, Viktor began to construct radio sets to his own original design. His dream was to become a theoretical physicist, and by the time he graduated from high school, Glushkov was already familiar with the basics of higher mathematics and quantum mechanics, which was usually only studied by senior students in universities. However, the Great Patriotic War interrupted his studies. The fall of 1941 brought another tragedy – the Nazis shot Glushkov’s mother. After Shakhta was liberated, Glushkov was conscripted, and participated in the reconstruction of Donbass coal mines.

In 1943, he entered the heat engineering department of Novocherkassy industrial institute. During his fourth year of study, which introduced specialized courses, Glushkov realized that heat engineering did not satisfy him as a major, and transferred to Rostov institute. In order to do so, he took all his fourth-year exams (covering all four years of physics and mathematics courses) ahead of time, and entered Rostov University as a fifth-year student. In 1947, he defended his thesis, where he presented a new method for calculating improper integral tables. Glushkov was assigned to the Ural, where he was supposed to work at one of the new establishments of the fledgling nuclear industry. His wife, whom he met on his third year at the university, went with him. However, as soon as he arrived at his new workplace, he received a new job appointment, this time to the Novocherkassy industrial institute. Glushkov and his wife had no money and could not return right away, so Viktor Mikhailovich found a temporary job at the Nizhniy Tahil pedagogical college, and later went to Sverdlovsk to look for work opportunities at the local institutes. Professor Sergey Nikolayevich Chernikov, who taught at Sverdlovsk University, helped young Glushkov to get an assistant position at the Forestry Engineering College.

In 1949, Glushkov began his graduate studies at the extramural department of Sverdlovsk University. By the end of 1950, he had already completed his dissertation “Theory of torsion-free locally nilpotent groups with the condition of cut-off of some chains of subgroups”, which he defended with stellar success and was immediately appointed associate professor.

In 1951, Viktor Mikhailovich began exploring a new branch of science, the topological space theory. In 1955, Glushkov was the first mathematician in the world to solve the fifth Gilbert’s problem. This became the basis for his Doctor’s Dissertation, which he defended at the Moscow University. Gilbert’s problem is one of the top 23 hardest problems in mathematics. Every new solution of the problem became a sensation in the world of science.

With the solution of the fifth generalized Gilbert’s problem, Glushkov entered the ranks of world’s leading algebraists. It seemed at the time that this would define his future career. However, the 1950s also saw the rise of Soviet electronic computer technologies. In 1951, a research team of the Academy of Sciences of the Ukrainian SSR headed by academician S.A. Lebedev created the first electronic computing machine in the USSR and continental Europe, the MESM. This project laid the foundation for future large-scale projects in the area.

This new branch of science drew Glushkov’s attention, and eventually the problems of computing technologies completely captured his imagination. This event determined Glushkov’s future scientific career. In 1956, he made the radical decision to give up his previous career and focus his attention on cybernetics, computing technologies, and applied mathematics.

From then on, Glushkov lived and worked in Kyiv. Here, he supervised the laboratory of computer technologies and mathematics of the Institute of Mathematics of the Academy of Sciences of Ukraine, which was created by Lebedev and already famous for its pioneering research in computing, and the creation of MESM and SESM machines.

At first, the laboratory staff approached their new boss with cautious mistrust, despite his likable personality. The “veteran” members of the staff, who used to work under Lebedev himself, doubted the ability of Glushkov as a theoretical mathematician to comprehend the practical nature of computer design. However, Glushkov quickly dispersed their doubts. He immediately familiarized himself with all the ongoing projects and began to actively participate in each and every one of them. For example, the design project for a two-machine system for radio location of air targets and fighter plane guidance was one of the major projects, on which the Institute worked prior to the new director’s arrival. Glushkov encouraged the creation of a strict scientific base for the project, and the formal expression of the mathematical theory for the guidance process. The ordering agency approved the results and implemented the technology in creating standard AA defense systems. Thus, none of the ongoing projects were abandoned. On the contrary, all of them were carried to a logical completion. According to Glushkov’s contemporaries, he had a talent for comprehending others’ ideas, and developing them further if they were worthwhile. Glushkov himself commented that a good supervisor does not get in the way of his ambitious staff members, while an excellent one steps in to assist.

In 1957, V.M. Glushkov became the director of the Computing center of the Academy of Sciences of the Ukrainian SSR; this organization was also treated as a scientific research institute (SRI). Five years later, the center gave birth to the Institute of Cybernetics.

In 1959, Viktor Mikhailovich began working on an interesting new project on automation of the robots’ motor function. He wanted to create an automated hand attached to a cart that would move along any control panel and flip switches, turn dials, etc. In addition, the machine would have primitive perception, the ability to recognize the position of dials or readings of scales. Unfortunately, Glushkov was unable to find an engineer interested in working with mechanical hands. He later commented that if the Institute of Cybernetics had good workshops, it would be the first in the world to create a mechanic hand.

Simultaneously, the Institute began working on creating a machine with the ability to recognize Russian language phrases, i.e. semantic networks. Glushkov himself developed the algorithms for this project. The algorithm built a semantic network based on the flow of input sentences, determining the relation of the words in a sentence. For example, “a chair stands on the ceiling” is correct grammatically, but incorrect semantically, and so on. The Institute also created a basic worldview model and came up with efficient coding. However, this project’s progress also came to a halt; it had to be tied to automatic translation, but the Institute once again lacked the necessary staff. Still, the idea was not completely wasted. In 1961, when Glushkov presented it at the Munich International Federation of Information on Processing (IFIP) Congress, it became a sensation. The Americans didn’t have anything similar at the time.

The computer “Kiev”, which was the first implementation of these ideas, became the first system in Europe capable of digital image processing and intellectual process modeling. It was connected to two original peripheral devices, which allowed the computer to model simple learning algorithms for image recognition and goal-oriented behavior, i.e. an image input device (from paper or film) and an image output device. At that time, the first output devices, prototypes for modern monitors, were a rarity and available only in the USA. In Europe, the Kyiv device was unique. In 1950s-1960s, “Kiev” was used for a series of studies on artificial intelligence, like teaching the machine to recognize simple geometric shapes, modeling reading machines for handwritten and printed text, tracking object movement in a series of pictures or a video, modeling group behavior in an evolutionary process, automatic preparation of computer function circuits, etc.

During an interview for the “Literaturnaya gazeta”, Viktor Mikhailovich commented, “There is no longer any doubt whether it is possible to create a completely artificial intelligence. It is definitely possible. It will not only be comparable to the human intelligence, but likely far superior to it. It should happen even before the beginning of the 21st century…”

In 1961, Glushkov published his famous monograph “Synthesis of digital automata”, which was later translated into English and published in the USA and other countries. Another of his key theoretic works, “Abstract theory of automata” was published the same year in the “Successes of mathematics sciences” journal. The article created a base for projects on the automata theory using algebraic methods. Many mathematicians-algebraists were inspired by this work and began researching this branch of science. In 1964, Glushkov received the Lenin Prize for this series of articles.

The importance of these works is difficult to overestimate. The use of the term “automaton” as a mathematical abstraction of structures an processes inside a computing machine opened completely new possibilities in computer engineering. Modern computer design automation systems wholly embrace these ideas.

Further development of computer design methodology required new technologies, specifically modular design methods. Glushkov outlined these fundamentals of computer design in a series of articles published in the “Kibernetika” (“Cybernetics”) journal in 1965-1966, and in the Soviet Academy of Sciences Herald in 1967. It was clear that effective computer use in the design process necessitated a comprehensive solution of all tasks in the process. This approach was used during the construction of the third generation of computers.

Due to the almost immediate switch to the fourth generation of computers (early 70s), Glushkov, along with Y.V. Kapitonova and A.A. Letichevskiy had to combine the process of hardware and software design.

The Institute of Cybernetics, based on Glushkov’s theoretical works, created a language for describing algorithms and computer structures, which were implemented in a unique series of systems, “PROEKT” (“PROEKT-1”, “PROEKT-ES”, “PROEKT-MIM”, and “PROEKT-MVK”). The experimental “PROEKT-1” machine was based on the M-220 computer and completed in 1970. The more powerful “PROEKT-2” was based on a two-machine system including M-220 and BESM-6 with an advanced system of peripheral devices. The overall size of “PROEKT-2” was 2 million machine commands. It was a suballocated specialized hardware and software suite with its own operating system and a specialized programming system. It also included the world’s first automation of the algorithm design stage. Glushkov’s team also developed a new method of complex program design, the formalized technical tasks method. Eventually, “PROEKT” systems were included in the unified computer system, and used as a prototype for the SAPR and SAPR BIS computers, which were used in many Soviet organizations.

“Automation of the computing machine design process”, a monograph by Glushkov, Kapitonova, and Letichevskiy, which summed up the experience of the “PROEKT” systems, was published in 1975, and earned the scientists the State Prize of the USSR.

In 1958, Glushkov proposed the idea of creating a multipurpose control computer. The project was realized within three years, supervised by Glushkov and B.N. Malinovsky (also Head Designer of the machine).

The switch from specialized control computers of the first generation (vacuum tube-based) to the semiconductor multipurpose control computers was vital for industry organization and allowed their use as automatic manufacturing process control systems.

The American multipurpose control computer project was initiated slightly earlier than the Soviet one, and completed in 1961, almost simultaneously with the launch of the Soviet multipurpose control computer “Dnepr”.

The Institute of Cybernetics also pursued another line of research, on creating computers for engineering calculations. The first machines in this line were the “Promin” (English: Ray) computers, serialized by the Severodonetsk instrument-building plant. It was an innovation of worldwide importance, and introduced a number of novel technologies. Most importantly, it was the first general-use machine with the so-called multistage microprogrammed control, authored by Glushkov.

The same control system was used again later in a series of engineering calculation machines, MIR-1, MIR-2, and MIR-3. These machines differed from others in that they had hardware implementation of machine language, similar to the high-level programming languages. The series of “MIR” computers interpreted the Algol-like “Analytic” language, developed by the Institute’s V.M. Glushkov, A.A. Letichevskiy, Y.V. Blagoveshchenskiy, and A.A. Dorodnitsyna. The team responsible for the creation of MIR-1 was later awarded the State Prize of the USSR.

In late 1960s, the Institute under Glushkov’s supervision began designing the “Ukraine” computer, which became a next step in intellectualization of computers, and in developing the architecture of high-performance universal computers different from von Neumann’s. However, “Ukraine” was never completed due to the lack of the necessary hardware technology at the time.

Still, the ideas presented in the “Ukraine” project were many years ahead of what the American scientists used in their universal computers in the 1970s.

In 1974, Glushkov presented a report on fractal computer design at the IFIP congress (co-authored by V.A. Myasnikov, M.B. Ignatiev, and V.A. Torgashov). In this report, he expressed his opinion that only a fundamentally new, non-von-Neumann architecture of computing systems could solve the problems of creating supercomputers, which could increase in productivity indefinitely through additional hardware. The idea of a fractal computer, supported by powerful recursive function base was also ahead of its time, but was never implemented because the necessary hardware did not exist.

This IFIP congress awarded Glushkov a special prize, the “silver core”. This was a way to recognize his considerable contribution to the organization as a member of the Program committee of the 1965 and 1968 congresses, and as the Chairman of the Program committee of the 1971 congress.

In late 1970s, V.M. Glushkov proposed the macro-pipeline computer architecture with multiple command and data flows (MIMD architecture under modern classification) as a way to implement non-von-Neumann architecture.

Once again, the Institute of cybernetics took up the task of designing the machine, this time with a team of Glushkov, S.B. Pogrebinskiy (head designer), V.S. Mikhalevich, A.A. Letichevskiy, and I.N. Molchanov. The two resulting machines, ES-2701 and the computing system ES-1766 were serialized by the Penza computer plant in 1984 and 1987 respectively. These were the most powerful Soviet machines at the time, with productivity exceeding 1 billion operations per second. The multi-processor system ensured an almost linear increase in productivity relative to hardware resources addition, and allowed for dynamic reconfiguring. These machines were completely unique and had no foreign counterparts. However, Glushkov never had a chance to see them in operation.

Glushkov was a world-famous authority in cybernetics. He based his understanding of cybernetics as a discipline on the works of A.I. Berg, A.A. Lyapunov, S.L. Sobolev, and I.A. Poletaev. He wrote numerous scientific articles on various issues in cybernetics, which were published in Soviet, British and American journals and encyclopedias.

His “Introduction to cybernetics” study, published in 1964, presented cybernetics as a science about general patterns, principles, and methods of data processing and complex system control. Computing technologies were seen as the primary instruments of cybernetics.

This understanding of the science was presented in world’s first “Encyclopedia of cybernetics”, prepared by Glushkov and published in 1974. Over 100 scientists from all over the USSR contributed to the encyclopedia, including more than 50 specialists from the Institute of cybernetics. In 1978, the team of its writers and editors received the State Prize of Ukraine.

Glushkov’s “life’s work”, however, was the creation of the Nation-wide Automated Economics Control System (Russian: OGAS). Unfortunately, this idea was never fully implemented, despite its potential to solve all of the economic problems of the Soviet Union.

In fall 1981, Glushkov’s health quickly deteriorated. He was first treated at the “Feofania” clinical hospital in Kyiv, and later sent to the Moscow Central clinical hospital. In January 1982, while being treated in the critical care ward, he asked his daughter Olga to record several stories about his life, summarizing his career. These tapes are preserved by the Glushkov family as an heirloom.

Viktor Mikhailovich died on January 20, 1982, at 58. He was buried in Kyiv, at the Baykove graveyard.

V.M. Glushkov published over 800 scientific works, more than 500 of which he wrote himself, co-authoring the rest.

Here is how another prominent Ukrainian scientist, Boris Paton described Glushkov’s contribution to science. He said, “Most people consider the task of creating and developing new major branches of science feasible only for large research teams, given the scale and complexity of the tasks. Nevertheless, when one mentally envisions the progress of this or that sphere of knowledge, it becomes clear that it is not faceless. More often than not, one can identify the connection between scientific achievements and the ideas of a prominent scientist or manager, who became their inspiration. Viktor Mikhailovich Glushkov was one such person; his name is inseparable from the creation of the cybernetic industry in our country.”