Hrachya Esayevich Hovsepyan in 1960-1976 was the chief designer of the Nairi computer family, which played an important role in the development of Soviet computer technology. He now lives in Los Angeles and regrets very much that he was forced to complete the work when there were unique prospects ahead. In an interview with the Museum project DataArt, Hrachya Yesayevich recalls how the first machines were developed from scratch at YerNIIMM, and explains why Nairi-4 did not become the first personal computer.
In the photo Hrachya Hovsepyan during a seminar at YerNIIMM in 1970
Start
I was born in Lebanon on June 12, 1933, and my family moved to Armenia in 1946. A very difficult time for the USSR, especially for Soviet Armenia. It was difficult to find bread, we were almost starving, but we survived. In 1949 I entered the Faculty of Physics and Mathematics at Yerevan University. In 1954 he graduated from it, worked as a teacher in the countryside for a year, and for another year at the Institute of Physics. There he gained experience in electronics and radio engineering.
Hrachya Hovsepyan is doing his graduation work in radio engineering at the Yerevan University, 1954.
Once I was present in the hall of the Academy of Sciences of the Armenian SSR, where famous Armenian scientists discussed cybernetics. Academician Iosif'yan talked about the science of management, which quite recently under Stalin was called pseudoscience. I also learned there that the young scientist Mergelyan would be in charge of the new Institute of Mathematical Machines. Impressed by this conference, I really wanted to go to work at the Mergelyan Institute. I had some kind of inner confidence that in this area of ââscience I could achieve some success. But it turned out that it is extremely difficult for a repatriate, that is, a person who has arrived from abroad, to get a job in a closed "mailbox". In the corridor, I accidentally met a familiar professor and asked him for help. The deputy director of the institute said at the reception that he could only take me as a laboratory assistant. I had to agreealthough I have already worked with electronics and could claim more. In general, then, thanks to the authority of Mergelyan in the Yerevan Research Institute of Mathematical Machines, it was possible to recruit young guys who had some experience in this area, mainly from Moscow.
"Hrazdan"
At YerNIIMM then the development of the semiconductor computer "Hrazdan" began. I was very interested in semiconductors, and therefore wanted to get into the laboratory of Efim Brusilovsky, who was appointed chief designer of the machine. The chief engineer opposed my transfer, referring to the fact that the group is fully formed. But Brusilovsky, apparently, liked me - he made sure that I was accepted.
Valentina Nazarova - Head of the Scientific Department of the Computing Center of the Voronezh State University - at the Hrazdan-2 computer console. The VSU Computing Center received the car from Yerevan in 1965 and used it until 1971. Photo from the archives of the VC VSU
There was very little information about semiconductors at that time, and we had to invent everything ourselves from scratch. I was entrusted with the trigger circuit, since I had little experience with an oscilloscope, transistors, resistance, etc. But theoretically I knew absolutely nothing, so I began the development by reading a book by the French author Vasseur. As a result, Brusilovsky liked my first trigger, and he appointed me the head of the group dealing with the Hrazdan control device. For him, I developed the entire schematic, and I got the original driver circuit. As a result, the delay was four times less than with the standard scheme. This was my first invention, and we named it the "Hrazdan" machine-saving scheme. She later became such for the Nairi car. All in all, we worked on Hrazdan for four years - until December 1960.
. . . 1956
Having entered the correspondence postgraduate study, my friend and I left Mergelyan and entered the Computing Center of the Academy of Sciences of Armenia. From there we went to Moscow - to the Institute of Automation and Telemechanics, where we became the graduate students of Professor Lerner. But a little later, from YerNIIMM we were urged to return, and the deputy director of the institute promised us that after the completion of the work on the implementation of Hrazdan, he would give us a more responsible job. The deputy director was Gurgen Markarovich Sargsyan - a very smart and good organizer. When Mergelyan returned to theoretical science, Sargsyan replaced him and confidently kept the institute on its feet.
Hrachya Hovsepyan during the November demonstration in Yerevan in 1957
In 1961, when the introduction of Hrazdan at the plant was completed, I was assigned a laboratory of small machines and was appointed chief designer. First of all, an order came from the Ministry of Radio Industry to develop a small electronic adding machine that could replace a large mechanical one. Since I had already read about the ideas of the English scientist Maurice Wilkes, I immediately decided: I will make a normal machine with microprogram control.
Maurice Wilkes' book was translated and published in the USSR back in 1953, just two years after the original edition. But amid the persecution of cybernetics in the mid-1950s,
In 1962, at the International Exhibition of Computer Science in Moscow, a French small machine CAB-500 of sequential action was shown. An ultra-modern, but very large and complex magnetic drum was used as a memory for it. The Ministry of Radio Industry bought this car, and we were ordered to repeat it. But our technologies did not allow this in principle, and my idea was completely different. I wanted to build a parallel action machine with microprogram control, and therefore I quietly worked on this business - I wrote a preliminary project. Then he showed it to very intelligent guys from the Computing Center of the USSR Academy of Sciences, who worked under the leadership of Academician Dorodnitsyn. They liked the idea very much - in fact, from it was born "Nairi".
CAB 500 SEA (Société d'Electronique appliquée à l'Automatisme)
«»: 1 2
When I started implementing it, many doubted that the idea would work. I myself did not know whether it would work or not, but I took the risk. Not imagining the possibilities of microprogramming, step by step I discovered its secrets. I went to Moscow to see a guy who was preparing a dissertation on microprogramming, and found other specialists. But I could not learn anything particularly valuable from others, I had to do everything myself.
I was fortunate again - there were now read-only memory devices on removable cassettes, each with a capacity of 2048 36-bit words. Without it, I would not have been able to implement the project: the machine had to be parallel action with a microprogram control principle, and programs and microprograms had to be stored in a single permanent memory. This was the main idea.
The parallel principle made it possible to use only one universal register-adder, and in the role of the remaining registers, fixed addresses of RAM were used. This significantly reduced the amount of equipment and reduced the complexity of machine installation.
It is very important that I was surrounded by a wonderful team: the laboratory staff - 26 people - understood the logic of my reasoning and believed me. Having started development in 1961, in 1964 we were able to present "Nairi-1" - a small machine with an original structure. It has been patented in four countries: England, France, Germany and Japan.
Although outwardly "Nairi" is close to the CAB-500, their structure and architecture are fundamentally different. This is confirmed by French patent No. 1.470.483 dated March 2, 1966
Our ministry did not approve the project at first, and we worked on it in secret. Only when the car was almost ready did the officials realize that something fundamentally new was being obtained. The State Commission for the acceptance of the machine was headed by Academician Anatoly Dorodnitsyn, who, after successful tests, gave the go-ahead for its serial production. In January 1965, in just a week of hard work, we completed the implementation of Nairi at the Kazan plant. They brought the project to crystal clearness. By the way, in Kazan "Nairi-1" went even earlier than at our pilot plant in YerNIIMM.
The ministry really wanted to be in time for the 800th anniversary exhibition in Leipzig, so it had to hurry. But we had time and presented the car with great success. American, German, French developers came to see how it was possible: all the machines are sequential, and ours are parallel, while smaller in size - about a desk. Then the Soviet trade representatives demonstrated the âNairiâ series in different cities of the world twenty more times. In particular, in 1976 in Los Angeles, local Armenians with great pleasure specially came to see "Nairi-4".
General view of the computer "Nairi-1", 1964. Serial sample of the Astrakhan computer plant
Finishing work on "Nairi-1", we were simultaneously working on the schematic of "Nairi-2". Even then I imagined how without much difficulty you can double the speed of the machine and the amount of memory, expand the software. So we have implemented Nairi-2 quite easily at our pilot plant and several other production facilities in the Union.
"Nairi-3"
We have already developed Nairi-3 on integrated circuits that we made in Zelenograd for space and military projects. But just the manufacturers wanted to bring this technology into wider use, and we could help them with this. The chief engineer of the Zelenograd mailbox, Lazarev, knew how widely Nairi-1 and Nairi-2 were used, the initiative to make a new machine on integrated circuits came from him.
It turned out to be a major breakthrough: our new development is two-stage microprogramming. One part of the ROM contains 300 bits of high-speed, but small amount of memory. In fact, they were placed in one cassette, and their addresses were memorized in another part of the ROM - already of a larger volume. That is, instead of memorizing 300 bits, we memorized only their binary addresses - 8 bits. In addition, we, for example, used an additional trigger, and in general we came up with a lot of things.
This made it possible to create a microprogram field - a virtually endless set of microprograms. Whereas IBM used 16-bit microprogramming, its methods severely limited the possibilities. We have got a huge memory - up to 128 thousand addresses with high speed. This opened up great prospects: we were among the first in the world to be able to apply methods of full firmware emulation of several machine languages. Thus, we could use the rich base of other machines, for example, the Minsk-22. For this, a modification "Nairi-3-1" was created.
Booklet about the computer "Nairi-3" for the international exhibition in Amsterdam 1971
The Nairi-3-2 modification was intended to implement a time-sharing system. Educational institutions and large factories were to become customers for it. One such machine could actually replace 128 "Nairi-2" if 128 teletype devices were connected to it. In the Soviet Union, our development was implemented in many areas: I saw how on "Nairi-3-2" they worked at the department of Academician Buslenko at the Gubkin Chemical Institute. Each student, sitting at a teletype, as if received for work "Nairi-2". Of course, this provided high efficiency.
Also, on the basis of "Nairi-3-2", a two-machine complex "Nairi-3-3" was developed to automate processes at large industries. It, for example, was introduced at the Lviv Radio Components Plant. Of course, the military was very interested in "Nairi-3": they were attracted by the flexible structure and easy adaptation of the machine to solve special tasks.
An interesting moment in this history is the continuity of two ideas that played a significant role in the history of computers. The first was expressed by John von Neumann in 1945: he proposed storing numbers and commands in one common memory. The second - to store a number, commands and micro-commands in the common memory - we proposed in 1966, developing "Nairi-3".
Fragment of an article by Academician Ershov about the features of the Nairi-3 computer in the American specialized journal Datamation for September 1975
Cheating the minister
There was a moment when they wanted to remove me from the development of "Nairi-3". The car refused to work at the pilot plant, and the new director of YerNIIMM Fadey Tachatovich Sargsyan claimed that I had failed everything. Math winged me at a party meeting, despite the fact that I have never been a member of the communist party. In fact, it was like that.
"Nairi-3" is the first car of the third generation, it was very interested in the military-industrial complex. As soon as we finished it constructively and started adjusting it, the military - they are in a hurry - called and demanded to take the car to a secret electronics exhibition in Moscow. We also went with the whole team: at night we were engaged in adjustment. But during the day, the lights on the control panel shone as if everything was already working. In general, window dressing. Suddenly, the minister of radio-electronic industry Kalmykov came up to me: âHrachya Yesayevich! Now a member of the Politburo Ustinov and the head of the military-industrial complex Smirnov will come to check the car. Please tell me that it is already being implemented at the Astrakhan plant. " Wanted me to lie. In ten minutes they really come. Ustinov says: âI see the machine is good, third generation, based on integrated circuits, and I am accused of being backward ... Comrade Kalmykov,why is it never mentioned anywhere? " Kalmykov answers: âWhat are you talking about! This machine is already being implemented. The manufacturing process is underway at the Astrakhan plant. " âYou pleasantly surprised me,â Ustinov rejoiced and left.
Dmitry Ustinov was one of the main organizers and ideologists of the Soviet military-industrial complex. In the photo he is on the podium during a meeting of the Politburo of the CPSU Central Committee. Late 1970s
When I returned to Armenia, I saw that either ten or 20 people from the Astrakhan plant had already arrived. They took all the documents, although we had not finished yet: the chief engineer of the experimental plant Ishin himself dreamed of becoming the chief designer of the machine. He also claimed that three Nairi-3 computers were in their setup department, but they could not work because of development errors. In fact, the reason was non-compliance with the technology in their manufacture. I knew that with my help everything would work out in Astrakhan, but just at that moment they tried to remove me from my post.
The residents of Astrakhan demanded support, they were told that they needed to finalize the documentation. But I said, "Let me go, people are waiting." Since the management did not believe that I would succeed, the business trip was allowed. In Astrakhan, I organized work in three shifts: I arrived at 7 am, left at midnight. Ishin called the director of the plant and demanded that I - a deceiver - not be allowed to work. But he talked to me and did not interfere. Three months later, "Nairi-3" was ready: seven vehicles were put on the conveyor. So the ministerâs deception helped me - accelerated implementation and saved me from collapse. Otherwise I would have been eaten.
When the production was launched, Sargsyan came to Astrakhan and began to scold Ishin: "He deceived me, bastard, he informed me incorrectly." Asks: "How can I help you?" "Nothing, I answer, has already been done." Still, Fadey played unfairly then. But when we received the State Prize for "Nairi-3", he began to treat me well.
Then I studied Nairi-4. If not for the circumstances, I would have done Nairi-5. But I was not lucky, circumstances did not allow.
Conclusion of Academician Dorodnitsyn's review on the development of the Nairi computer family in connection with the awarding of the State Prize to the authors of the development
"Nairi-4"
It was already a complex of universal computing tools (UMC), since the entire computing system was divided into parts, like cubes, from which you can then assemble a machine of the desired configuration. This idea could be realized through a unique microprogramming method. Large and super-large integrated circuits were used as the element base, which made it possible to reduce the size of the machine.
A truncated processor became the main component of Nairi-4. It consisted of a microprogram control unit containing a read-only memory for storing microinstructions, microprograms, machine instructions, programs, subroutines, diagnostic tests, etc. Then an arithmetic unit, a control unit for external devices. In the minimum configuration, Nairi-4 consisted of a truncated processor and a typewriter connected to it as a device for communicating with the machine. But the use of the Unibus bus made it possible to easily expand its capabilities to the desired configuration, connect additional devices: I / O, external storage, etc.
KUVS also allowed to create systems of control machines and time sharing systems, moreover, "Nairi-4" was software compatible with PDP-11. No other machine provided such opportunities. We managed to plan the development of holographic memory for storing microprograms, we even prepared a technical assignment jointly with Vasily Bukreev from the Moscow Center for Physical Instrumentation. That is, already for "Nairi-4" we were going to use a CD.
I must say that I was pleasantly surprised when, in the TSB section dedicated to Soviet computers, I accidentally saw my name next to the machines of the Nairi series. The rest of the machines were listed without the names of the developers.
Operator at the control panel "Nairi-K" - modifications of "Nairi-1" with increased RAM
Hard choice
1976 was the most unexpected and unpleasant moment in my life. When the repatriates were allowed to leave the Soviet Union, many Armenians started to leave abroad, and my family - brothers, sister and mother - also wanted to leave. I was faced with a difficult dilemma. The OVIR knew that I was a famous scientist and that I worked at YerNIIMM. For the family to be released, I had to give up my career.
It was the biggest tragedy of my life. I received the State Prize, communicated on equal terms with ministers, made reports for many institutions on the development of computer technology. Just at this time, the head of the head office of the radio industry, Gorshkov, called me to Moscow to tell the directors of the enterprises subordinate to him about Nairi-4, and then said: "Now, guys, go and work like YerNIIMM." And against the backdrop of such a career, I need to decide whether to continue working. As a result, in order not to leave my family in a miserable situation, I gave up my career. I still cannot forgive myself for taking this step.
Missed opportunities
Still, I think Fadey Tachatovich Sargsyan did not like at all when ministers invited me to Moscow. And the rapid growth of my authority was unpleasant to him. Therefore, when I voiced the decision to leave him, he, knowing his true reason, did not call me up for a conversation or look for ways to solve the problem. In my opinion, he, on the contrary, was delighted. Gathered a banquet for 50 people, made a speech about what I did, what kind of people I prepared. He emphasized that the work will not stop - there is someone to replace me. I spent it "worthily".
In the mid-1970s, we missed the option of making Nairi-4 for consumer goods, that is, we never made a domestic personal computer. The truncated Nairi-4 processor in the TV case was supposed to be a model for him, the model had already been made and stood for a long time in the director's office. In fact, everything was ready in terms of ideology, structure, and software.
But after my departure, the Nairi-4 was taken away by the military, and the further development of the vehicle went only in the field of military use with a rigid structure. In this regard, one can recall that Fadey Sargsyan came to YerNIIMM as a military representative, and later received the rank of General of the Army.
They say that "Nairi-4" is the beginning of personal computers in the USSR, but we developed it before IBM. Already at "Nairi-3-2" with time sharing, any student could work as on a personal computer. Teletype or display - what's the difference! The idea is the same, but the system itself in the Union was already in operation in 1976. It is a pity that the scientists of that time did not widely advertise it. A lot has been done, but the development has remained in the shadows.
Refusenik
Formally, I explained the decision to leave YerNIIMM by the fact that the climate of Armenia does not suit my Muscovite wife. We moved to Moscow. Everything was fine at first. I worked both at the university and at the Academy of Sciences. But then my brother did a disservice: when he arrived in the United States, he immediately sent me a challenge without asking me. When the KGB found out about the documents, I was asked to quit my job, and I was no longer accepted into scientific organizations. It can be said as an enemy of the people. I didnât work for three years - wherever I turned, they refused. I tried to explain that I have a lot of experience, that I can be useful, but no. It happened under Andropov.
The family of G. E. Hovsepyan during the depressive years in Moscow after being refused to travel abroad, 1984
Once, when I was walking in the street, the police caught me. When they found out that I didnât work, they took me to the department and said that they would be imprisoned for parasitism if I didnât get a job in a month. I didn't want to sign the protocol, but they forced me to. He told about this to an acquaintance guy who had already come this way, he explained that a job really needs to be found urgently. He also suggested that I become a stoker in the boiler room. I came to the basement of one of the high-rise buildings in Moscow, where 16 people were sitting around the table. Half were former criminals, the rest were refuseniks, Jews of various professions. One of them confirmed that he knew the Nairi car. The boss said that no one is interested in political issues here and asked: "Can you make us such a tester so that he could check the health of the control unit, which turns on the pumps in series?" âI can do anything,â I replied.
In a basement full of water and shit, I worked for five years - until 1988. Then my twin sons grew up, papers came from the military registration and enlistment office. I went to the OVIR to find out when our family will finally be released. I was told that in the next five years I definitely will not go anywhere.
In protest, I went on a hunger strike: "Either release me, or I will commit suicide." He wrote applications to all addresses - including foreign embassies. On the 16th day of the hunger strike, I could no longer work, I went to the clinic. The doctors confirmed that I was really in poor condition, but they didnât give me sick leave, because it was my own fault.
Hrachya Hovsepyan after a hunger strike in Moscow, 1988
Then I turned to the American embassy. US Ambassador Jack Matlock said he had no right to provide me with medical assistance and asked to stop the hunger strike. But I kept on fighting. On the 30th day, a medical worker in a dressing gown arrived from the KGB and said that if I didnât stop, they would take me to the hospital tomorrow. I guessed that nothing good awaited me there. At 12 o'clock in the morning, an Armenian guy connected with the KGB called with threats: "You will be very bad if you don't stop." I thought and decided not to risk it.
Later, I was often invited to the US Embassy: I met with congressmen, representatives of the Jewish community came from America. One woman said they made a big splash in Washington on my behalf. And then President Reagan came to Moscow, for this event an American correspondent interviewed me and my relatives in Los Angeles. The plot was shown on TV in America.
The very same story from Hrachya Hovsepyan's interview to NBC TV The
question was resolved at a meeting between Reagan and Gorbachev: in a week I and about ten other refuseniks received permits and left for America. But in the USA I was not allowed to work in my specialty either. They didnât trust me: what if Iâm a KGB agent? I had to become an adjuster in one not very successful Vietnamese personal computer repair company. I worked for three years and left. This is how my working career ended.