Only a handful of people know that in November 1953, half a year after Lebedev and his team completed BESM, the first sequential computer, TsEM-1 [in Russian: Tsifrovaya Elektronnaya Mashina-1], went on-line at the Institute of Atomic Energy in Moscow and operated until 1960. The decision to develop this machine was almost accidental. At that time, Sergey Sobolev was Kurchatov's assistant director at the Institute of Atomic Energy, and in 1950 he happened to come across the description of the ENIAC in an American magazine. Being aware of the Strela and BESM projects, Sobolev handed the American magazine to the supervisor of the institute's measurements laboratory, N.A. Yavlinsky. The magazine then turned up in the hands of a young specialist, Gennady Alexandrovich Mikhailov, who had graduated from the Ivanovsk Power Institute just three years prior. Among the scarcely available foreign publications, he was able to find only a couple more articles in British journals about the EDSAC computer, which was constructed at the Cambridge University. Unfortunately, these journals presented only the flowchart and operational features of the machine. The binary system, as well as programming, was not widely known at that time, and there were no textbooks on solving problems using numerical methods. There was yet another difficulty: the team that designed and assembled the TsEM-1 consisted of only four people – two engineers and two technicians – Mikhailov included.

Just like the MESM and BESM's designs are attributed exclusively to Lebedev, the TsEM-1's scheme belongs entirely to Mikhailov's.

The TsEM-1 contained an operating memory of 128 binary 31-bit digits on 32 mercury delay lines; each one had 16 digits with a sequential retrieval rate of 512 kilobytes per second. The memory capacity was later extended to 496 digits – 4096 digits on a magnetic drum. Data input and output were performed using as ST-35 telegraph apparatus. Digital print-outs on telegraph tape were copied onto 5-track perforated tape, and data input from the same perforated tape was sent through a photographic reader at high speed. The machine's operational modes were observable on an oscilloscope – a precursor of our modern digital displays. The average addition and subtraction speed of the TsEM-1 was 495 operations per second, 232 operations per second for multiplication and division. It contained 1900 electronic vacuum tubes, consuming roughly 14 kilowatts of power per hour. The machine was housed in six metal racks measuring 80 x 180 x 40 centimeters each. The main physical limitations of the TsEM-1 were found in its mercury delay lines: because of its 1000 millimeter long, 18 millimeter diameter quartz acoustic radiator, it was necessary to constantly check for sharply focused ultrasonic rays and for the reflection levels from the receiving quartz. Luckily, weekly preventive maintenance guaranteed consistently reliable operation of the TsEM-1.

Like MESM and BESM, TsEM-1 was an original project, based on ingenuity and imagination of its creators, and it was substantially different from EDSAC. For example, multiplication was carried out by rounding off; division was done without recovery of a remainder; and a two-address instruction system replaced the previous one-address unit. Lebedev proposed these improvements during the TsEM-1's construction. In addition, the command modification system by means of "control characters" was unique. It facilitated program compression, which in view of the computer's limited immediate access memory was very important.

Even within the Atomic Energy Institute TsEM-1 was not acknowledged in its early years. The supervisor of one the institute's branches, physicist Lev Andreevich Artsimovich, was initially quite skeptical about this kind of technology. After some time, he changed his mind and found the computer to be useful and powerful when he saw what it could produce: at the end of 1954 Mikhailov had programmed and solved an equation on the TsEM-1 that described the process of plasma filament compression in experiments on controlled nuclear fusion. S.M. Osovtsev, who was a member of the theoretical physics team headed by Mikhail Alexandrovich Leontovich, had set up the equation. At first, Artsimovich rejected the result of the accelerated plasma filament compression with oscillations overlaid on it. However, after three or four days of theoretical analysis, he obtained the same results. A great number of calculations on nuclear reactors functions and dosimeters were made on the TsEM-1; Lebedev, M.D. Millionshchikov and others became quite familiar with the machine. Mikhailov adds some new touches to the portrait of Lebedev:

In the 1950s, working as a staff engineer at the Kurchatov Atomic Energy Institute, I was fortunate to meet many of our distinguished scientists. Some of them I only saw from afar at lectures and seminars, people like Kurchatov, Kikoin, Tamm, Ioffe, Timofeev-Resovskii, and Sakharov. Others, such as Sobolev, Artsimovich and Leontovich, I developed closer, more personal relationships with.^[13]

When I defended my Master's thesis, the test administrators for computer technology were academicians Artsimovich and Lebedev. It still gives me a great pleasure to think about being in the company of those two brilliant scientists, plus remembering many other talented scientists of the 1950s and 1960s. My only worry is that if Sergei Alexeevich were to be judged solely by his appearance, he would have looked ordinary compared to his colleagues; he had neither a remarkable statue nor a determined face. But it was his humility along with his immeasurable talent which made Sergei Alexeevich stand out above the rest.

I heard of him for the first time from my lab colleagues, who referred to him as an exceptionally talented scientist. Our team, headed by N.A. Yavlinsky, moved to the Nuclear Power Institute where Lebedev was working. Yavlinsky and Lebedev were friends and their families spend a lot of time together until 1962, when Yavlinskii perished in an airplane crash along with his wife and son. Thanks to that friendship, I had the pleasure of seeing Sergei Alexeevich at family parties as well. Even then, he remained unobtrusive and plain, without a hint of self-flattery or false modesty.

In 1959, Mikhailov moved to Kiev and became a department head at the Ukraine Academy of Sciences Computer Center. He continues:

Summer of 1961 was the last time Sergei Alexeevich visited Kiev, which had always been dear to him. He visited our computer center that had already moved from Feofania to Lisogorsk. We organized a trip to Feofania so Sergei Alexeevich could once more see the place where he started his work. By that time, he had achieved almost everything: he had become an Academician, a Lenin prize-winner, a Hero of Socialist Labor... it seemed to be the time for honors. But Sergei Alexeevich was rather modest and would never allow grand meetings, banquets, or celebrations to be held for his arrival. There was no secret about his coming, but only a few of us knew about it.

Once, at his anniversary celebration at the Institute for Precision Mechanics' conference hall, he looked very embarrassed and uncomfortable dressed in an Uzbek robe and tyubeteika [an embroidered Central Asian skullcap], while great fuss was made over the entire procedure.

I never heard a bad word about him. But, at the same time, it would not be true to call him an infinitely kind soul. On the aforesaid Masters' examinations, Sergei Alexeevich quietly gave a deserved "two" (equivalent to an "F") to his own graduate assistant. I remember when we discussed defending the theses, he made an ironic remark: 'In our institute we have a division of labor: some people make machines, others defend dissertations.'

Having visited our laboratory and scrupulously tested TsEM-1, Sergei Alexeevich surprised us with this question: 'Don't you bang it with a hammer?' It turned out that a rubber mallet was a common laboratory tool used on the BESM, and banging it on the machine's solid-state metal frame was typical machine maintenance! No less surprising would be an order not to work on a problem longer then fifteen minutes, unless it required recalculation, so as not to waste the machine's time.

Everything mentioned above relates to the first generation of computers with vacuum tubes. Second-generation machines were developed without them, and Lebedev's first semiconductor machines were the BESM-3 and BESM-4. These machines emerged as a result of youthful enthusiasm: these developments came about at SKB and at the Institute for Precision Mechanics outside of the state "plan" – at the initiative of the young engineers and technicians.

Lebedev's team member, A.A. Gryzlov, recalled that a relatively small group of young coworkers – engineers, technicians and self-taught inventors – was commissioned to master the first semiconductor components in 1964 in order to prepare the SKB staff for the upcoming BESM-6 project. First, they were given the task of developing prototypes for the computer's main units to gain some design experience. The final prototype was named BESM-3M. Inspired by their success, the young coworkers proposed a daring idea: to develop a new machine based on the available prototype which would repeat the block-logic diagram of the M-20 computer, only using new components. SKB's head at that time, O.P Vasiliev, supported the young workers' ideas, and Lebedev raised no objections to the "unproven" youth and thus, BESM-4 was born in the creative and friendly atmosphere that pervaded Lebedev's institute.

The State Committee headed by Dorodnitsyn noted the high performance and innovative features of the first Soviet multi-purpose semiconductor computer. The machine was characterized by its reliability, small size, and low price; it quickly became popular among users. When a BESM-4 was installed in the Soviet Academy of Sciences Computing Center and people inquired about it, the always answer was: "Your machine demoralizes young engineers. They don't have to do routine maintenance checks because the machine is error-free. It is too reliable." Nothing else needed to be said.

^[13] Editor's note: Mikhailov is referring to the famous Soviet physicists Igor Tamm, Isaak Kikoyin and Abram Ioffe, and biologist Nikolai Timofeev-Risovskii.