Vladimir Vorobyov: “Mikhail Botvinnik was happy that the car did not find the solution he found behind the board”

When the Research Computing Center was established in Leningrad in 1974 (now SPIIRAS: St. Petersburg Institute of Informatics and Automation of the Russian Academy of Sciences), Vladimir Vorob'ev entered the computer software supply group. In 1978, a laboratory of the same name was formed at its base, and Vladimir Ivanovich supervised it for 30 years.

Vorobiev - Doctor of Physical and Mathematical Sciences, Professor. He now holds the position of chief scientific officer, and his place of work is called the Laboratory of Information Computing Systems and Programming Technologies.
')
We continue to complement the DataArt museum collection with interviews on IT history. Earlier, Sergei Zonov , Yevgeny Kanevsky and Alexander Lamden became our interlocutors.

About the weather

- When did you realize that computing is yours?

- Still a student. In 1983 I graduated from the Hydrometeorological Institute, we had an experimental group of programmers to calculate the weather forecast. I wrote my diploma work at the Novosibirsk Computing Center, which was headed by the last president of the USSR Academy of Sciences, Gury Ivanovich Marchuk. In his group, I developed a model of the atmospheric boundary layer.


The Institute of Computational Mathematics and Mathematical Geophysics (ICM & MG) was founded in 1964 as the Computing Center of the Siberian Branch of the Academy of Sciences of the USSR

- Why precisely the border?

- Here the biggest problems are turbulent flows, which are very difficult to describe. When I was involved in this work, the errors in the predictions of the boundary layer were the biggest - about 200 percent. Because it did not take into account the interaction with the underlying surface, and it is very heterogeneous: water, forests, arable land ... This significantly affects the weather.

- The program you wrote helped to reduce the percentage of errors?

- We brought it to reasonable numbers, about 70-60 percent. Now the prediction of the prediction (not an error, but justification) - 80-90 percent. It all depends on what we predict. If the temperature, then it is very variable in space and time, you need to choose representative points for measurements.

- How was the weather forecast in the 1960s?

- Forecasters manually processed prognostic maps. Now it makes the computer. Without computing technology it is impossible to master the gigantic amount of information received from satellites and with the help of observations. This is a serious job. There is a huge network of stations in the world - there are thousands of them. Part - automatic. Data from them and from satellites are flown to regional centers, and then transmitted to the World Meteorological Organization. Information accumulates in world weather centers. One of them is in Obninsk.

Further, the data are received by regional organizations for operational work. There are compiled synoptic maps, where all weather information is marked in a special code. With the help of this map and the weather forecast map that the computer calculates, the forecaster makes a decision and finally formulates the forecast - a maximum of three days.

- That is, long-term forecasts are unreliable?

- This is a theoretically proven fact. The atmosphere and the ocean are interrelated, without them we can not consider the weather. The system is non-linear, after three days we lose data on the initial state.

CYBER 170

- How did you get back to Leningrad after Novosibirsk?

- After graduation, I entered graduate school, which at the time was not easy to do. It was required to work out, but I managed to convince the management. In parallel, I continued to work in Novosibirsk - I flew there on business trips. Having defended my thesis in 1969, I became a member of the Institute of Oceanology of the Academy of Sciences. There he did the same plus a number of other tasks. Since space missions had already begun, I calculated the radiation on Mars, the solar climates of the planets of the solar system.

- What is the practical significance of this research?
- The weather on Earth is completely dependent on the energy that comes to us from the sun. When we compare different planets, we better understand the physical processes taking place in us. In addition, the institute was engaged in the general circulation of the ocean atmosphere. It was one of the main topics.


Central racks BESM-6. On the left are the buffer registers and the communication device, then the control unit (control unit), then the control unit (arithmetic unit). On the right is the edge of the HCS (control of external devices).

- Why did you go to the Computing Center in 1975?

- There were problems in relations with the management, in addition, I was offered a promotion. If I worked there as a junior researcher (even doctors of science were in this position), here - immediately as a senior. Career growth.


Machine room: tape drives, foreground (yellow) teletype

- What was your responsibility now?

- Partially I was engaged in modeling, but the main task was the mathematical software of computers. The main workhorse then was BESM-6. It carried out the majority of scientific research in the USSR. We had two such machines. In 1978, they bought the American CYBER 170 from CDC. Very expensive - about 5 million in gold paid for it. This is more than 5 million dollars. In total, there were 6-7 such cars in the USSR.


Magnetic drums external memory. The weight of the drum is about 300 kg, the memory capacity is 32 kiloslova of the BESM-6.

It was supplied with the software of the manufacturer - Control Data Corporation. It was much better than for BESM-6. Nothing superfluous, everything is very balanced, with powerful libraries of scientific programs that made it possible to facilitate the solution of many scientific problems.


The machine hall of the Computing Center of the Academy of Sciences on the Mendeleev line in Leningrad. Photo source

- What else is the American car superior domestic?

- In speed, efficiency and convenience of service. For example, she had a ready-made algorithm for solving differential equations. We only supplied the initial data and integrated the equation. At BESM-6, it had to be programmed manually.
In some industries - geophysics, the calculation of nuclear reactors - from computing equipment required high performance. In CYBER 170 it was higher. On Sakhalin, these machines were for the study of mineral resources.


CYBER 170 in the Computing Center. Photo of the late 1980s. Photo source

- What are the problems in the work machines had to face?

- When you start to solve a problem, in 90 percent of cases it does not go. Very rarely, programmers prepare an algorithm that is suitable immediately. We helped them find mistakes. Most often the machine does not understand the programming language when its rules are violated. There are software errors, sometimes equipment fails, for example, when the temperature rises. The car had three cooling cycles: air, freon and water. In the courtyard stood cooling tower. Have you seen these at power plants, where does steam come from? We also had it, only more modest. So, at the entrance the temperature should be 17 degrees. Sometimes when the heat or poor performance of the equipment, it increased, and the car crashed.

Starovoytova, Botvinnik and cosmonauts

- How many computing centers were in the USSR in those years?

- The first is in Moscow. This is the Dorodnitsyn Institute (Dorodnitsyn Computing Center of the Russian Academy of Sciences - Editor's Note.). There was a center in Novosibirsk - Guriy Ivanovich Marchuk was invited there. Their computer centers were created in serious institutions. For example, the Institute of Applied Mathematics of the Academy of Sciences, where calculations of the trajectories of spacecraft were performed. The hydrometeorological center has a computing center.
Since the machines are expensive and highly qualified specialists are needed for maintenance, not every institute could afford them. Therefore, they began to create centers for collective use. Now, by the way, we are also seeing a system of collective use, but at the new stage of development it is cloud computing.


Cyber ​​170 machine room. Operator works behind the terminal.

- How many specialists worked in your computer center?

- About ten electronics engineers, about the same number of software specialists. Plus, refrigerators, electricians ... This is for the whole fleet of machines that Leningrad institutions of the Academy of Sciences could use.

- What tasks came to you?

- We found out the needs of institutions. The Institute of Oceanology has been modeling the ocean, atmosphere and climate. Physical and Technical Institute - Solid State Physics. He had his own cars, but we were interested in libraries of scientific programs. By the way, on CYBER 170, not all tasks could be considered, because the Americans controlled our calculations. And in those days there were also sanctions. We were given a car with a capacity of no more than that of BESM-6. But we removed the jumpers, lowering the speed.

- There were cases when someone tried to get to you without a queue?

- Yes. For example, Galina Starovoytova. She came to me and insisted on counting her task. She said: "The director has allowed me." Galina Vasilievna then worked in the Kunstkamera and was engaged in some kind of sociological research. To solve its problem, an algorithmist was needed to translate all this into the language of mathematics, and then to find a numerical algorithm that would fit the architecture of a specific computer. The task was overwhelming.

- I had to refuse?

- Yes, even though it was not easy. Galina Vasilyevna demanded persistently and energetically. But we were ready to provide only advice on solving the problem.


Cassette Recorder Cyber ​​170

- Did bribes offer you to count out of turn?

- Somehow I noticed that the user service group has cakes - they stand outside the window. It turned out that people are pushing their tasks, because machine time was a big deficit. I had to stop and send to work in this group of his deputy. For such an unscientific task, he even took offense at me.

In general, maintaining order in the center was not easy. From the subscriber points work was carried out at night. I had to do computer security. Some users, especially physicists, could penetrate where they should not have been - in the field of a system programmer, or in the resource of another user.

- Have any of the great people come to your center?

- For example, the sixth world chess champion Mikhail Botvinnik. We had a chess program, for those times it was rather serious. Mikhail Moiseevich uploaded his game with Capablanca and was very happy that the car did not find the solution that he found behind the board.


The party against Capablanca at the AVRO tournament in 1938 is the most famous in the career of Mikhail Botivnnik. After the 29th move, Garry Kasparov called the diagram "the most famous in the history of chess"

On excursions astronauts came to us. The connection with them was quite close, since our two first directors, Valentin Mikhailovich Ponomarev and Rafael Midhatovich Yusupov, came from the Mozhaysky Academy, which trains specialists for the space forces. In addition, our employees are still doing calculations for the space industry.

- What are you most proud of in your work?

- Creation of computer networks. This was our key task; we started it in 1985. Now everyone uses network services, the Internet is a common thing, and at that time we conducted experiments. This is pretty serious work. What is a network tie? Here we are conducting a dialogue, there is a certain protocol of the interview. You ask a question, I answer. Computers also conduct dialogue over the protocol. The first sends a request, the second confirms: “Yes, I am ready to receive!” After that, a packet is formed in which there is an address - where it goes and to whom.

- Who used computer networks in those years?

- The same meteorologists, for example. They had networks between weak machines that collected data from stations, accumulated them, and transmitted them to large computers. The launch of the spacecraft also did not do without computer networks.