First computer

It is obvious that the first successful device in this area were the accounts that were used at least 5 thousand years ago and are successfully used to this day. From the history of technology are also known all sorts of computational mechanisms, primarily functional analogs of modern arithmetic calculators. The first mechanical calculator was made in 1623 by German Wilhelm Schickard. There were even more devices that performed specialized computational functions. But calculators or specialized calculators can be called computers only with many reservations. What is different from them is the PC, which is on your table? Of course, versatility. Modern computer is freely programmable to solve an extremely wide range of tasks. In theory, he can solve any problem and model any process that mathematics can describe. This property has allowed computers to drastically change the entire civilization of the Earth for several decades. At first glance, it seems that humankind has gradually come to this type of computer and it makes no sense to look for the “same” first universal freely programmable computer. Some people spread this opinion to most inventions and discoveries. But such an opinion is wrong. Computer revolution began in the 30s - 40s of the XX century.

In 1939-41, John Vincent Atanasoff and Clifford Berry built (but not finished) an ABC computer (Atanasoff Berry Computer) to solve systems of algebraic equations with 30 unknowns for the money of a small American company Research Corporation. The device had a binary number system and a memory of 1632 paper capacitors.

In 1944, Howard Aiken (Howard Hathaway Aiken - the only American founding father of computers who was not called John) and 4 more engineers under a contract with IBM built a machine for calculating ballistic tables, known as Mark I. Its main elements were gear wheels (to represent numbers) and electromechanical relays (to control the computation process), it weighed 5 tons and occupied an area of ​​several tens of m2 at Harvard University. The car had 72 registers, each of which was a device of 24 gears and had a mechanism to transfer dozens to another register. 23 wheels served to represent the number, one for its sign. Separately for constants was provided a mechanical memory of 60 registers. For the operations of multiplication and division, as well as to calculate the sine, natural logarithm and exponent, separate computational units were used. Harvard Mark I worked on a program that he read from a perforated tape. In short, it was an advanced adding machine, replacing the labor of about 20 operators with conventional manual adding machines. Some researchers consider Mark I the first really working computer.

However, the majority of textbooks and encyclopedias call the lamp-relay ENIAC (Electronic Numerical Integrator and Calculator) as the first "real" computer, which was launched in December 1945 at the University of Pennsylvania. Designers - John Presper Eckert (John Presper Eckert) and John Mouchli (John Mauchly) - were thoroughly familiar with Atanasoff's device and borrowed some of his ideas. ENIAC also had a lot in common with Mark I, but it worked about three orders of magnitude faster. As it turned out relatively recently, in the UK in June 1944, a modified version of the Colossus computing device began to work, functionally in no way inferior to ENIAC, ENIAC had a much more modest size. These machines solved narrow problems: ENIAC was intended (like Mark I) for calculating ballistic tables, Colossus for searching ciphers in coded messages. The cost of building and operating both computers was astronomical and only the military could afford them. Nevertheless, in terms of their architecture (design and operation principles), Mark I, ENIAC and Colossus had little in common with modern computers, they were rather giant programmable calculators. In ENIAC, 10 vacuum triodes were connected into a ring, forming a decimal counter (which served as the counting wheel of a mechanical machine), 10 such rings plus 2 triggers for representing the sign formed a memory register. In total, ENIAC had 20 registers, each of which was the operators set the sequence of operations provided with a transmission scheme of tens and could be used for the operations of summation and subtraction. Other arithmetic operations were performed in specialized blocks. The numbers were transferred from one part of the machine to another through 11 conductors, one for each decimal place and one for the sign. The value of the transmitted digit was equal to the number of pulses that passed through this conductor. The operation of individual units of the machine was controlled by a master oscillator, which determined the sequence of clock and synchronizing pulses, these pulses "opened" and "closed" the corresponding electronic units of the machine. Entering numbers into the machine was done using punched cards, and the sequence of operations was set using Kurbels and switching fields, as in the PBX. The size of ENIAC has long been the talk of the town. He weighed 30 tons, contained 18 thousand vacuum tubes, 70 thousand resistors, 10 thousand capacitors, 7200 diodes, 1500 relays, and a dial pad comprised 6000 switches. It occupied an area of ​​about 200 m2 and consumed a power of 160 kW.

In August 1944, mathematician John von Neumann joined the group under the strictest secrecy of developing ENIAC. Less than a year later, von Neumann (whom we incorrectly call Neumann) prepared a 100-page report on the work plan for the promising EDVAC machine. The document was, of course, secret and was called "First Draft Report on the EDVAC". To the dismay of Mouchley and Eckert, a certain Goldstein, who was present at the report as a military representative, duplicated a copy of "First Draft ..." and, under the authorship of Neumann, sent it to several scientists in the United States and Great Britain. The report became widely known and turned into the first classic work on designing computers. The ideas of the creators of ENIAC, their predecessor John Vincent Atanasoff, and even Alan Turing were outlined in it by von Neumann (the history of science knows a lot of such incidents). Mouchly and Eckert, as typical Americans, filed a lawsuit against Neumann, but in the end they did not achieve the truth and they had to leave the Defense Ministry and establish their own company. The EDVAC project, as I understand it, was implemented in 1950.

Thus, for a long time it was John Von Neumann who was considered the founder of modern computer architecture. The computer described in the famous report had:

1. Special organs for performing the simplest arithmetic functions (addition, subtraction, division and multiplication).
2. Logical control was carried out by one central control unit.
3. The computer had a large amount of memory in order to execute long sequences of commands.
4. Bodies to transfer information from external media to the central arithmetic part, to the controlling unit, as well as to memory (that is, in modern terms, the input device).
5. Bodies to transfer information from the arithmetic, controlling parts and memory to an external recording device (i.e., output device).

The basis of his work was based on the following principles:

1. the principle of program management, according to which the program consists of a set of sequentially executed commands;
2. the principle of uniformity of memory, when the program and data for it are stored in a single storage device (RAM, in modern terms;
3. the principle of targeting, according to which (operational) memory consists of numbered cells, at any moment accessible for reading (by the processor).

The machine had to work only with integer mathematics, and the input and output of data would occur directly through the arithmetic unit, and not through the bus, like in modern computers. Nevertheless, it is believed that these theoretical prerequisites formed the basis for the further development of computer technology in the US and the UK.

But back in 1934, on the other side of the barricades, in Reich III, a 23-year-old student at the Berlin Polytechnic Konrad Zuse (Konrad Zuse) invented a new device, whose architecture and operating principles in general coincided with modern digital computers. His device had (then still theoretically) a control unit, a calculator (combining arithmetic and logical operations, that is, a processor) and memory. Zuse then believed that the computer should be based on the following six principles:

• software management ;
• binary number system ;
• floating point arithmetic;
• fully automatic arithmetic calculations;
• large capacity memory;
• yes / no elements.

It seems to me that the main thing is that Zuse was the first to understand that the bit must be the basis of computer data processing (he called it “yes / no status”). This means that any calculations can be made based on elements (like a relay) that have two physical states (closed and open). Konrad Zuse also introduced the concept of conditional propositions for formulas of binary algebra and invented the “machine word”.

Shortly after graduating from the polytechnic, Zuse joined Henschel, where he was involved in aerodynamic calculations. Obviously, this circumstance stimulated his work on computers. He decided to make a programmable device that works with binary numbers, in which the control unit and the processor are separated from the Z3 memory unit. In 1936, he made a mechanical storage device based on binary elements (moving metal slats), and received a patent for it. In the same year, in a small room in the apartment of his parents, Zuse began to build his first computer, which he named V-1 (Versuchsmodell-1, V-1). In 1938, work on the V-1 (renamed Z1 by then) was completed. It was an experimental or demonstration model, unable to solve serious practical problems due to the small amount of memory and unreliable mechanical processor. Nevertheless, the Z1 allowed Conrad to get a position and support at the German Aviation Research Institute. Using the same memory, Zuse by April 1939 built the following computer model (Z2), which had a processor on electromechanical relays. After this success, the designer was taken to the army for a year. After serving, he returned to college. At that time, the used relays were available to Zuse in large quantities and he decides to assemble a serious car with them, with the same architecture as the Z1. This car - Z3 - was officially "handed over" on December 5, 1941, and the author received a patent for it. Apparently, as a bonus, Zuse was again taken to the army and sent to the front.

Z3 was the first universal freely programmable digital computer with an ideology that is still in use today. On the Z3, the Z3 structural diagram can be clearly seen that this computer was surprisingly close to modern. The binary memory contained 64 22-bit floating point numbers (comma) and was connected to the processor (arithmetic unit) by a data bus that transmitted the numbers and the mantissa separately. The processor that handled the floating-point binary numbers was connected by bus to the decimal I / O devices: a four-button keyboard and a lamp panel. The control unit contained a circuit for each command and synchronized the operation of all components. The clock frequency was approximately 5.3 Hertz. The program was packed on punched tape, which was a film, using nine 8-bit commands (input, output, read from memory, write to memory, square root and four arithmetic operations). In practice, the Z3 team did not allow to implement the conditional transition and this is considered today the main drawback of this machine. Nevertheless, theoretically (this is shown in the works of modern researchers), the versatility of the Z3 was limited only by the amount of memory for storing data.

The manufacture of Z3 took about 2600 relays, including 1800 for memory and 600 for the processor. The machine consumed about 4 kW of power. At that time, she (like all of Tsuze’s cars) could be considered portable, she weighed about a ton and her dimensions were ten times smaller than English and American ones. It should be noted that Zuse did not use vacuum tubes in his machines as elements only because of a lack of free space and insufficient funding.

Recalled a few months from the front, Zuse decided to create a more powerful and sophisticated computer. Well aware that the main thing is a large amount of RAM, Zuse decided that it should have a capacity of at least 1024 bits. It was assumed that the new computer will be equipped with 2 perforators and 6 readers punched tape (including subroutines), as well as an automatic printing device. The computer should also have an extended set of commands that would allow conditional jumps and address translation. Z4 Z4 managed to build and run to the very end of the war. By the time the bombing was destroyed Z3. Because of the difficult military situation, the Z4 had to be transported from place to place. April 28 in an underground structure in the mountains of the Harz Zuse showed his leading German aerodynamics (among which were Ludwig Prandtl and Albert Betz). In the end, the Z4 was only saved thanks to the staff of Werner von Braun, who hid him in a shed in one of the alpine villages. Actually, the Z4 had a processor of 2200 relays, a mechanical memory of 64 32 bit words (memory was planned for 500 words), two devices for punching / reading punched tape, a decimal keyboard, an output device in the form of an electric Mercedes typewriter. He worked at a frequency of 30 Hertz, and weighed and consumed energy approximately like Z3. In a sense, it was a personal computer, since its maintenance was simple and, most importantly, it was easily programmed by one person. To program a Z4, it took about three hours to solve a typical problem.

Drawing nodes Z1

Konrad Zuse himself denied all his life that his computers were used in Nazi Germany for practical calculations. But it is hard to believe for two reasons. First, its only surviving computer - the Z4 - after undergoing several minor modifications after the war, was installed at the Institute of Applied Mathematics in Zurich (Eidgenoessische Technische Hochschule), where it worked almost realistically for five years (it was one of two computers working in Europe at the time, the second was MESM Sergey Lebedev). Then he was transported to France, where he worked about the same. Currently, the Z4 can be seen at the Deutsche Museum Munich. Secondly, immediately after the war, almost all the specialists who collaborated with the Reich government were turned into intellectual slaves: the majority of them were dismantled by the victorious countries, forced to work in their military departments, many were brought to trial and imprisoned, while those who remained at large in Germany lived in constant fear. It was to the last Zuse and belonged. Immediately after the war, he was arrested, but Konrad obviously managed to "otmazatsya".

Since 1942, Zuse hatched the idea of ​​an algorithmic programming language. Today, any educated person understands that a programming language is as important as the hardware of a computer. A few months after the end of the war, Zuse developed an algorithmic language for engineering calculations. He called it Plankalkul (u with two dots, of course). Plankalkul introduced the concept of an object; it allowed working with subarrays of data, subroutines, and even arrays of programs. Zuse invented an assignment operator and defined a separate character for it. In terms of its level, Plankalkul corresponded to the ALGOL 60/68 language, which was common in the 1960-1970s.