Nikolay Nikolaevich Govorun
Corresponding member of the AS USSR
Edition and translation: Alexander Nitusov
Nikolay Nikolaevich Govorun was born on the 18 th of March 1930 in Ukraine, in the settlement Shevchenko (Andrianopol) near Voroshilovgrad. His father was a factory worker.
Nikolay Nikolaevich Govorun
In 1953 N.N. Govorun graduated from the faculty of physics at the Moscow State University – MSU. During the two following years he was working as engineer-designer at the transporting machinery construction plant in Kharkov (Ukraine), but in 1955 returned to Moscow to continue his study at the MSU as post-graduate.
That time he entered the chair of mathematics at the faculty of physics. Famous mathematicians A.N. Tikhonov and A.A. Samarskiy became his scientific supervisors. They not only developed his mathematical talents but also brought him up as a scientist of new type, who could personally conduct researches and at the same time organise and lead teams within fundamental scientific programs. The programs, which could be realised only by collective efforts and subsequently served as basis for applied scientific problems.
As a post graduate Govorun prepared five scientific publications on integral equations in theory of antennae. During research he obtained essential experience in programming: he had to solve problems of electromagnetic waves diffraction on bodies of revolution with high conductivity. The programs were written in machine codes of computer STRELA, what needed detailed understanding of both the physical process and programming technologies. All that gives a solid reason to say that, besides physics and mathematics, Govorun obtained the third qualification – programming.
He finished studying in 1958 and received a position at the Laboratory of Theoretical Physics of the Joint Institute for Nuclear Research (JINR) in academic town Dubna (125 km North from Moscow). He quickly realised that it was exactly the place where computers should be most actively implemented into physical researches. Therefore, already in 1959 scientists of JINR presented results of (computer) analysis of the films, which were shot in the bubble chamber of the particles tracking detector, on beams of a synchrocyclotron. Programs of the analysis were prepared on computer URAL by Govorun and his colleagues. Later he authored many original algorithms and programs for processing photos from the bubble-chamber detectors, spark-chamber detectors and Wilson chamber detectors (basic tools for experimental researches in nuclear physics). Most of the programs were written for computers M-20, BESM-4, Minsk-22. Those computers were modified at the JINR by Govorun and his colleagues to be directly connected with measuring instrumentation and used in on-line mode. Initially they were not intended for such purposes.
In 1961 he submitted dissertation thesis “Integral equations of the Antennae Theory” and received the first scientific degree.
At the JINR Govorun also founded, and subsequently always developed, a big library of programs for processing the spectrometric information, obtained during experimental researches in a branch called ‘particle physics', or ‘high energy physics'. It started with selection of programs for BESM-6 and ES-computers prepared by collaborators of the Laboratory of Computing Techniques and Automation, established at the JINR in 1966. First, Govorun was appointed vice-director of the laboratory and later, since 1988, its director. Besides the JINR own software the library was always collecting programs provided by many other –Russian and foreign- institutions and various research centres. Among them there were Moscow State University, Leningrad Institute of Nuclear Physics, Moscow Institute of Physics and Engineering, Institute of Physics in Tbilisi ( Georgia ) and many universities and institutes of the countries – members of the joint nuclear researches. Fruitful international co-operation in that field was possible primarily due to Govorun's personal efforts. Its materials were widely used both in the USSR/Russia and abroad (more than 20 of -only big- research centres). In 1986 its leading collaborators – N.N. Govorun, V.P. Shirikov and some others, were awarded the “Premium of the USSR Council of Ministers”.
From 1967 till his early death in 1989 Govorun was the chairman of mathematical department of the Council on Automation of Scientific Researches at the AS USSR Presidium. The department intensively popularised “Collection of Software Libraries and Programming Complexes” (including the JINR library) and conducted organizational work in that field during more than 20 years. Numerous presentations were performed at international conferences, exhibitions, seminars, etc.
In the end of the 1960-s Govorun initiated development of a translator for BESM-6 from high-level programming language FORTRAN. That was the first realization of FORTRAN for high-performance computers, most popular in the USSR. Before that, only Algol-60 (as a high-level standard language) was in use. According to the popular, among many soviet scientists, opinion it was the only programming language able of meeting both numerous scientific and “common” demands. In reality, Algol-60 did possess more developed mathematical basis and its description was more rigorous. Therefore it was more attractive for pure mathematicians. However the JINR, being a partner of CERN, needed similar (at least compatible with the partner's software) selection of programs for processing of increasing amounts of data. As the computations of the European research centres were chiefly based on FORTRAN, Govorun performed series of “adopting” projects and eventually introduced it in the USSR as additional scientific tool and also as the language of communication between physicists and mathematicians, thus facilitating their exchange with algorithms. He also extended FORTRAN applications in -more general- programming of scientific and engineering problems. In reality, it wouldn't be big exaggeration to say that Govorun was the leader of the FORTRAN – applied-implementation in the USSR.
Govorun also participated in development of -general- software for BESM-6 operating system, on all levels. The computer was initially provided with operating system D68. In the end of the 1960-s JINR and the Moscow State University jointly developed one more operating system for BESM-6 - OS “Dubna”. Govorun and his colleagues proposed multi-language monitoring system “Dubna”, which maintained communications with extended libraries and also managed various tasks. The system integrated the mentioned FORTRAN translator, translators for autocode/machine-oriented language/ MADLEN, for ALGOL, etc. Intensive co-operation of the system software designers notably enriched software selection of BESM-6, which itself became a technical basis for an important stage in development of system programming in the USSR.
Govorun with his team efficiently solved problem of automated software supply for open heterogeneous measuring and computing complex, which consisted of various computers (BESM-6, CDC 6500, ES). In reality, that was rather a local network. Modular programming system “Gidra” (similar to “Hydra” at CERN) was designed for information retrieving from films, which registered physical experiments, and for its subsequent processing. In fact that was a library of functional modules. It generated FORTRAN texts of programs for experimental data processing. They were based on specifications, which were informing what kind of experiment had been performed, what facilities (including instrumentation) implemented and it also displayed characteristics of the utilised computers and used algorithms.
Programming modules of “Gidra” were independent and the data exchange was performed via shared memory area, what simplified procedures of programming elements correction and extension. The system essentially increased efficiency of programmers' labour - they were able to begin development of new software even before the arrival of experimental information. The programs themselves were more easily adapted to computers (and languages) of higher level.
In 1987-1989 Govorun organised development of software complexes for theoretical mathematical modelling of muon-catalysed fusion. That project was performed by an international of mathematicians and physicists. Govorun himself conducted precession calculations of multi-parameter models of muon-catalysis, what granted the JINR priority in that domain. He also initiated development of algorithms for parallel computations, including both the domain of framework models in quantum chromo-dynamics and experimental information processing in general. He maintained close co-operation with designers and producers of super-computers, in particular with academician V.A. Melnikov2 – designer of computer “Electronica-SS-BIS” (comparable with “Cray”).
As the JINR was important scientific institution it received all necessary facilities. However, the computers of various age produced by different plants (M-20, “ Kiev ”, Minsk-2, BESM-4, Minsk-22, etc.) were incompatible and could not be integrated into joint multi-machine complex, what seriously hampered progress of research methods. Neither special software nor appliances for their connection were available. Govorun managed to solve the problem by designing hierarchic computer complex (multi-machine) for automation of scientific researches, which was eventually produced in the JINR, in the 1970-s.
Lover level of the hierarchy integrated mini-computers, built-in control systems of experimental physical installations and also devices for experimental information retrieval and its preliminary processing. Its middle level consisted of various computers mounted in measuring and computing centres of the JINR laboratories. The highest level was formed of BESM-6, CDC-6500 with complex common terminal network and later extended with a cluster of five ES computers (ES-1037, -1060, -1061) with shared disc memory and with one more cluster of two computers VAX-8350. As the JINR maintained direct connection with CERN, each of the CERN computers in Geneva also was accessible for data exchange and joint solving of highly complicated problems. In the 1970-s, creation of so complex system was extremely difficult work, chiefly based on combining of incompatible computers, both on the hard- and software (system) levels. Nevertheless, the common institute terminal network JINET (J oint I nstitute NET work) was in regular operation already in 1985. Software for the (local) network was designed at the laboratory of computer engineering and automatics. In 1988 the net – JINET JINR – was integrated into international computer network via packet switching node (on protocol X.25) at the Moscow Scientific-Research Institute of Automated Systems. The institute technically supported on-line communications of the Soviet scientific centres with their foreign partners and other organisations (but that was not yet Internet).
In the 1970-1980 Govorun's personal experience and authority notably influenced development of bus-modular data processing systems and experiment automation based on international standards САМАС and Multibus-2.
In the middle of the 1980-s intensive usage of personal computers in scientific and administrative work of the institute caused obvious need in more progressive local network. It took him not much time to organise design of such net. Even more, he used the opportunity and launched another project – creation of global informational net, connecting Soviet centres of nuclear research with analogous foreign institutes. The net should utilise both satellite and cable communication channels. It was the KOKOS 3 project.
It is obvious that the role of N.N. Govorun in development of the Soviet school of system programming was very important. His contribution to progress of the scientific computer networks (local) is equally notable. Constantly growing level of the works of his followers and assistants was a good proof of that and it remains so in our time. He also paid attention to updating of his own scientific qualification. Thus in 1969 he successfully submitted dissertation thesis and was awarded a doctorate (“Doctor of physical and mathematical sciences”) – his second degree. In 1972 he was elected to the USSR Academy of Sciences as a corresponding member.
During several years Govorun was lecturing at the Moscow State University, on computer software development and data processing systems. He published about 200 scientific works and in 1977-1988 was editor in chief of the journal “Programming”, which efficiently combined fundamental theoretical problems with actual publications of applied character.
Govorun was decorated with two orders “Red Banner of Labour” and several medals.
He possessed many talents; his scientific heritage in development of information methods and technologies in physics is very rich and interesting, although it impresses with variety of subjects. Unfortunately his life was not a long one. N.N. Govorun died in Dubna on the 21 st of July 1989.
Based on (Russian) biographical article by E.N. Filinov and V.P. Shirikov.
1. JINR was established in town of Dubna, in the first half of the 1950-s. It became stable partner of European Centre of Physical Researches – CERN (Switzerland). In spite of its “atomic nature” it has been always a civil scientific organisation with higher international reputation. “Academic town” – scientific centre/institute situated well off a big city.
3. KOKOS – Russ. KOsmicheskaya KOmpyuternaya Svyas (Space Computer Communication)
- Programming, 1991, N 03, pp.5-30.
- “Nikolay Nikolaevich Govorun” (1930-1989). //Dubna, Joined Institute of Nuclear Researches (Bibliography of N.N. Govorun's scientific works). 1990.
- “Nikolay Nikolaevich Govorun. Book of memoirs”. V.P. Shirikov, E.M. Molchanov (Eds). Composed by A.G. Zaikina, T.A. Strizh. //
Dubna, Joined Institute of Nuclear Researches. 1999.