English Version

Aleksandridi Tamara
????(???), alexandridi@mail.ru

Automatic Digital Computer M-1

The scientific report “Automatic Digital Computer [M-1] was issued in December, 1951 at the Electrical System Laboratory of the Power Supply Institution attached to the Academy of Science of the USSR. The report was developed under Isaac S. Brook’s supervisor. The report was confirmed by the academician of the Academy of Science of the USSR G. Krzhizhanovsky on the 15th of December 1951. The report was the first USSR document on creation of a Russian computer. The computer underwent all the tests successfully and served both for the purposes of the Academy of Science of the USSR and other organizations.

I. Brook’s research on computers is dated back to the year 1948. He was the first in the USSR (together with B. Rameev) to develop the digital computer with hard program control. The first USSR certificate on the invention of “DC with a common bus” was received by them in December, 1948.

The Decree of the Presidium of Academy of Science of the USSR issued on the 22nd of April, 1950 said about the construction of M-1. I. Brook started to man the team. The first one to hire was Nickolai Matyukhin, the recent graduate of Moscow Energy Institute (College of Radio Engineering). I. Brook familiarized him with the main idea of the new digital computer. Working in close cooperation with each other, they worked out the architecture, design, main characteristics of the-would-be computer. Nickolai was an active participator in finding new design solutions, further on N. Matyukhin (under permanent support and guidance of I. Brook) practically functioned as the Chief Constructor of the project. 

 

Here is an extract of Nicolai Matyukhin’s memoirs:

“I’d like to render the atmosphere of the pioneer years in the field of computers.   

 

The team is formed and the ADC M-1 construction is started in the year of 1950.

Isaac Brook is hiring a team of young graduates from Moscow Energy Institute

 (College of Radio Engineering), there are seven of us:

Two junior scientific co-workers    -    A. Zalkind

                                                             N. Matyukhin

Two graduate students                          T. Alexandridi

                                                              M. Kartsev

Three technicians                              - Yu. Rogatchev

                                                                R. Shidlovsky

                                                                L. Zhurkin

My first task, assigned by Isaac Brook was to build a tube diode three-input adder which served as a test of my professional validity.

The second assignment was to design a typical desk top.

The third assignment was given to me when I had already been appointed as the head of the team. We started to build the ADC M-1.

While designing M-1, we faced a lot of challenges. One of them was nearly complete lack of the component details. I. Brook found a witty solution of this problem – he used the World War trophies stored in the country. As a result, the backbone of the project consisted of the following ideas and “trophies”: 

- combination of components of most various origin;

- just two types of electronic tubes;

- cupric rectifiers of gauges;

-magnet heads of a tape recorder;

- electron-emitting tubes of an oscillograph;

- trophy teletype which used to be operating in the Vermacht headquarters.

 

About the style of leadership of Isaac Brook:

- deep understanding of a goal, clear vision, and imaginative argumentation;

- empowerment and delegation of responsibilities;

- great team worker.”

 

Creating computer M-1, we had to handle various problems – beginning with the voltage regulator for powerful motor-generator of direct current, served as the source of secondary power, and finishing with the development of the system of commands and first tasks programming. We had to develop the architecture of the computer, its main devices and the system of components, facing the above-mentioned problems in the spare parts supply.

ADC M-1 consisted of arithmetic device, main programming data unit (control device), inner memory of two types – (fast – based on electrostatic tubes and slow – based on a magnet drum), input-output devices using a telegraph letter printer.

         In the fall of 1950 the N. Matyukhin’s team was manned.

The arithmetic device and the logic system were developed by N. Matyukhin and Yu. Rogatchev, main programming data unit was developed by M. Kartsev and R. Shidlovsky, , memory device on a magnet drum was developed by N. Matyukhin and L. Zhurkin, memory device on electrostatic tubes was developed by T. Alexandrididi, input-output devices was developed by A. Zalkind and D. Yermochekov, power supply device was developed by V. Belynsky, construction and design – by A. Kokalevsky. Complex computer testing and of programming testing technology were headed by N. Matyukhin. All the team players were very creative and inventive, using spare parts from other devices – magnet tape recorder heads, oscillograph electron-emitting tubes, letter printing teletype, etc

While designing and developing M-1, quite new technical solutions were offered, to wit:

two-address command system, which found wide use in the future in our country and abroad. N. Matyukhin wrote about this original technical solution: “ The choice itself of the command system was not easy for us – at that time the most common and most natural was considered to be the three-address system, which dated back to Newman’s works and which demanded quite huge ranking of registers and memory. Our limited choices motivated us for searching the most economic solutions. As it often happens, decision came quite by chance. Isaac Brook hired a young mathematician, Yu. Shreider. He was learning the ABC of programming together with us, and eventually, he paid attention to the fact that for many formulas for approximate calculations the result of the operation becomes one of the operands for the next step. This was the basic solution for the first two-address command system. Our solutions were approved by I. Brook, and ADC M-1 became the first two-address computer.”

In M-1 diode logic schemes were built on transistors. It was quite a revolution for digital computers at that time. Initially the tube diodes 6X6 were used in the model of a binary gauge. But, again, concurrence of circumstances helped. Trying to decrease the number of radio tubes in the computer resulted in an attempt to try cupric rectifiers KVMP-2-7 instead of radio tubes 6X6.

 

 Cupric rectifier KVMP-2-7 parameters:

              Allowed direct current                        4 Ma

              Direct resistance                             3-5 Kom

              Allowed back voltage                     120V

              Back resistance                     0,5 – 2 Mom 

             

             The ratio of direct and back resistance (not lower than 1: 100) provided for diode function performance in logic schemes. But one had to figure out the possibility to use them in impulse schemes at the quite high frequency.  Experiments held for testing logical schemes, developed with provision of the above-mentioned parameters, showed sufficient operation stability. Testing of a model of a digit of the arithmetic device with an adder, deciphers, and mixers based on KVMP-2-7 rectifiers in various modes and deviation levels of power supply devices considering and different parameters of spare parts of was conducted.  The model performance testing was successful. In August of 1950 the concluding testing was performed with direct participation of I. Brook. The testing performance endorsed reliability of logic schemes performance, based on cupric rectifier KVMP-2-7.

              Again, N. Matyukhin’s memoir: “One of the crucial decisions, which, to my mind, predetermined the success of our first computer and brief time of its creating, was Brook’s decision to use transistors. At that time they were presented in the industry only by small sized cupric rectifiers, which were being produced for the needs of measuring engineering. Brook made an agreement with a small enterprise on the production of a special modification of the rectifier of the type we needed, the size of a conventional resistor, and we developed the set of standard schemes. In a shop, attached to the laboratory, we began to produce and install the blocks of the computer, and less than in a year, the computer was “alive and breathing”, though it counted a number of thousands of tubes and a few hundreds cupric rectifiers. Thus, ADC M-1 became the first computer in the world made on transistors. 

       Great challenges were faced while designing electrostatic memory. It’s a well known fact that both in the USA and USSR special electron-emitting tubes – potential scopes - were used for these purposes. But in the USSR these tubes were designed only for the organizations that developed computers according to the government decrees. Unfortunately, we couldn’t count on these special tubes produced for our laboratory, because our research was planned as an initiative scientific work planned by the Academy of Sciences of the USSR. So Isaac Brook made a decision to use standard electron-emitting tubes LO-737, traditionally used in oscillogrpahs.

In September of 1950, Brook recruited a graduate student of Radio Engineering College of Moscow Energy Institute, Tamara Alexandrididi. Brook suggested she should accomplish a diploma research work on the topic: “Electrostatic Memory Device,” designed for ADC M-1.

First the research was carried out on a standard oscillograph with LO-737 tubes and supplementary measuring schemes.

Even the first testing showed that information memory effect on the screen of an electron-emitting tube might be put into practice. Immediately designing of electrostatic memory device had been started.

Let’s briefly describe the physical processes on which electrostatic memory (EM) is based. Information is stored in the memory as a definite set of static charges placed on the screen of a electron-emitting tube. Memory effect is based on the phenomenon of secondary electron emission. At the definite value of accelerating voltage, the coefficient of secondary emission is more than one, which means that, when bombarding the screen by the beam, the number of secondary electrons, leaving the screen, is more than the number of primary electrons, arriving at the screen. Therefore, the bombarded part of the screen becomes positively charged. For writing the binary information the system of reading-recording “focus-defocus”, at which “1” was recorded by a focused beam, and “0” – by a defocused beam, and reading was performed by a defocused beam. When reading “1” a positive signal appears, but the information is being razed. That’s why, regeneration is being performed after reading, that is “1” is being recorded again.

In the spring of 1951, under the academic supervision of I. Brook, T. Alexandrididi defended her diploma work on the topic “Electrostatic Memory Device.”

By that time EM had been already made at the laboratory, and its testing had begun.

            By the end of 1951 ADC M-1 had been put to pilot performance first time in the world. It included fast working memory device based on standard LO-737 electron-emitting tubes. 

    By design EM consisted out of 8 LO 737 tubes and scanning and control blocks.

On the screen of every tube there were 32 lines, and every line consisted of 25 points, that is one number or a command. So, the total volume of a fast working EM was equal to 256 25-digit binary words. EM operated in 50 mc sec, that is 20 000 operations in a minute was accomplished.

             In the fall of 1951 the tuning of M-1 computer had been completed. By December of the same year, the computer had undergone complex testing successfully and the industrial production of the computer was started.             

                 M-1 main parameters

                 Number system - binary.

                 Binary digit number  - 25.

                 Command system – two addresses.

                 Volume of inner memory:

                       on electrostatic tubes   - 256 addresses,

                       on a magnet drum - 256 addresses.

                Speed:

                       with slow memory - 20 op/sec

                       with fast memory adding operation was performed in 50 mc sec,

                        multiplying operation – in 2000 mc sec.

                Number of electron tubes - 730.   

                Power supply - 8 KW.

                Needed area - 4 square m.

     

         Prominent scientists arrived at the laboratory to get familiar with ADC M-1 operation including academicians A. Nesmeyanov, M. Lavrentyev, S. Sobolev, A. Berg. The academician S. S. Sobolev, Deputy for Research Work at the Kurchatov Institute then, was one of the first scientists who solved the problems on nuclear research    using M-1.

Computer M-1 had been operated for three years, and for a year and a half was the only one operating computer in Russian federation. Its architecture and many of principal scheme decisions served as a base for industrial production of computers M-3, “Minsk,” “Razdan,” etc.

Creators of computer M-1 became prominent specialists in the area of computer engineering. They contributed a lot in its development, heading various scientific, educational and industrial organizations. Their work was highly estimated by giving them scientific and honorary ranks and governmental awards. 

information
sponsors:
 
sponsors: