Hungary’s first operating digital programmable computer, the “MESZ I”, was completed in 1958. (The first operational electronic computer in Hungary was an M-3 vacuum tube computer somewhat redesigned and built in Budapest just a little bit after MESZ I)
Currently the MESZ I can be visited in a museum in Budapest (Technical Study Stores, The Hungarian Museum of Science, Technology and Transport). It is not operational, yet, fortunately it is in good condition. The reason for this is partially that its relays are rather worn out, and partially that the cables connecting the cabinets were removed when it was relocated to the museum.
The aim of our team is to rediscover and understand the operation of the computer in order to create a MESZ I emulator on a present-day device. It should be an easy task to write such a software, but unfortunately we do not know exactly how the machine functioned, as all the detailed documentation seems to be lost. Only a few overview articles have been found. The best one, which happens to be in English, is accessible here.
We are looking for people who were present at the construction and installation of the computer, who used it, or are acquainted with the applied technologies. We will continue to look for documentation and anything else that could be of use; however, the most expedient approach is to reverse-engineer the computer by tracing the cabling of the relays.
This is an enormous challenge that we do in our free time and thus progress is rather slow. Nevertheless, we already have some encouraging results. We dedicated this website to share these findings with those who are interested.
The Brief History of MESZ I
The MESZ I was designed and mostly built by Professor László Kozma (1902-1983), the head of the Department of Wired Telecommunications at the Budapest University of Technology, and later the dean of the Faculty of Electrical Engineering.
Between 1930 and 1942 he worked at the Bell Telephone Manufacturing Company in Antwerp, Belgium. Kozma was involved with the development of new telephone exchanges and with designing calculators from telephone exchange relays as well. 37 single and co-authored patents bear his name. Amidst the storms of WW2, Kozma returned to Hungary in 1942. After surviving the hardships he faced during the war due to his Jewish heritage, his troubles continued when the newly established communist regime in Hungary imprisoned him on fabricated charges. However, his talents were eventually appreciated and he was appointed as a head of department at the newly established Faculty of Electrical Engineering at the Budapest University of Technology in 1954.
Kozma began to design the computer around 1955. It was built during 1957-58 and was documented to be in operation by November of 1958. It was in use until around 1966, then, in early 1967, it was relocated to the museum where it is located to this day.
The computer was installed in the building “St” of the Budapest University of Technology. Next to the main entrance of the building a plaque and a relief commemorates the machine and its constructor. In addition, his pioneering work at the Bell Telephone Manufacturing Company and on the MESZ I was posthumously recognized with the prestigious IEEE Computer Pioneer Award in 1996.
The MESZ I acronym stands for “the first electronic calculator of the University of Technology” (I stands for the roman numeral 1). Although we call it a computer today, the word did not yet exist in Hungarian at the time.
The Application of the MESZ I
The MESZ I was built for educational purposes. Kozma’s aim was to demonstrate the operation of digital circuit design, which was already prevalent in the control of telephone exchanges.
The programs written for the MESZ I included a cubic equation solver, a schedule maker for the first division of the national soccer championship, and even a rather simple Russian-Hungarian translator. Although it was not primarily designed to, it was used in research and development projects as well–for example, it was used during the dimensioning of the Hungarian-made telephone exchanges that were installed in Cuba.
The computer was built mainly from telephone exchange parts, both because these were readily available and because of Kozma’s familiarity with them. Its main components were the relays, more precisely the Hungarian version of the American Western Electric’s “R” type flat type relays. The usage of relays fit the didactical purpose of the machine perfectly, as their operations are visible to the naked eye as well. In addition, the cabinets have see-through glass doors in order to make these observations possible. As the operation of the machine could be paused after each step, the internal workings of the computer could be followed throughout the execution of the program.
The MESZ I consists of four cabinets of relays, each containing 648 relays in 27 columns and 24 rows. Thus, the whole computer contains about 2600 relays altogether.
The first cabinet is the memory with the capacity to store twelve words of 33 bits each. In contemporary terminology, this is just below 50 bytes in the space of a wardrobe! Staying with the modern terminology, the second and third cabinets function as the processor (CPU). The fourth cabinet was made subsequently in 1960 to hold an additional 12 words and thus double the memory of the MESZ I.
The control desk was made out of a school desk. It contained the following: meters for the power supply, a program card reader, a data input unit, a “debugger” module, a simple program counter, and a printer as an output device.
The program card reader is a unique design, which, due to its appearance, we simply call “the waffle maker.” It can read one card at a time. The data input unit is a keyboard made out of telephone exchange parts, and a few push-button controllers. These were used to enter the input parameters during the program’s run, and probably to start the program as well. The debugger could be used to pause the program execution, and to run it one step at a time. It is connected to the desk with a long cable, so it could be brought all the way to the cabinets and follow the states and operations of the relays step-by-step. The program counter is a string of lights at the back end of the table. One can keep track of the program execution as every light represents one step in the program. Since a program card could hold 45 instructions, there are 45 lights in the control desk. Finally, the output device is essentially a rather old (even at the time) electric typewriter. In order to function as a printer, magnets were installed under numeric and some other important keys, which were controlled by the computer.
For images please turn to the Gallery page.
The narration of this clip in English is:
The students often have to solve difficult mathematical problems at the Faculty of Electrical Engineering of the Technical University. Prof. László Kozma describes the first Hungarian automatic calculator, designed by him, and manufactured using domestic parts. The instructions of the machine is assigned on a punched sheet [“decal”/”sticker” in original]. It solves several hours long computing tasks within minutes. The digits of the problem are entered via keys, the relays start to operate, and the small bulbs show the progress of the work. The typewriter on the control desk prints the result by itself. A quick check and it seems to be correct. The designer, the students, and even the machine can be satisfied.
- Word length: 33 bits: 27 bits for the mantissa and 6 bits for the exponent (characteristic)
- Random access memory (RAM): 2 ∙ 12 words (in today’s terms: 792 bits, or 99 bytes)
- Read-only memory (“hardwired ROM”): a few constant values (the exact number and the values are still unknown)
- Program memory (read-only): 45 instructions on an external punched card (12 bits per instruction) and further 15 ∙ 12 = 180 data bits per card
- Instruction set: 32 instructions, which most probably include:
- multiplication (and probably division as well)
- comparisons (less than, greater than, equal to)
- data movement
- conditional instructions (branching)
- Speed of execution: n/a
- Clock frequency: most probably no clock signal was used (asynchronous machine)
Balázsfi, Diána, PhD – Neurobiologist
Németh, Krisztián, PhD — Computer scientist
Nyilas, Gergely Sándor — Computer scientist
Vid, Gábor – Electrical engineer
We would like to thank Kendra Chilson, Gábor Képes, László Kiss and Máté Szabó for their help.
Reach us at: