Germany surrendered to the Allies on May 8, 1945, which is now called V-E Day and stands for Victory in Europe. Japan surrendered on August 15, 1945, V-J Day (Victory over Japan). These
surrenders ended the immediate hostilities between the Allies and the Axis powers, but they did not reduce the need for either analog or digital computers for military use.
Soon after the end of hostilities, tensions began to occur between the Soviet Union and its former allies. By 1947, this tension had started to be called “the Cold War” and this state of mutual hostility lasted until 1991.
A group of computer engineers meeting in New York at Columbia University on September 15, 1947, decided to form the Association of Computing Machinery (ACM), which is an important business association of computer manufacturers. This organization has become one of the largest technical associations in the world. Its original mission statement is still valid today: “The purpose of this organization would be to advance the science, development, and construction and the application of the new machinery for computing, reasoning, and other handling of information.” There are currently about 100,000 ACM members and 170 regional and local chapters. Because of the diversity of computer and software technologies, there are numerous special interest groups (SIGs) within the ACM umbrella. Currently, there are about thirty-five of these SIGs.
Many of the SIGs are famous in their own right, and their conferences are often venues where interesting new inventions surface. Among the special interest groups are SIGCOM for communications, SIGGRAPH for graphical topics, and SIGPLAN for programming languages.
There was a subcommittee on large-scale computing within the American Institute of Electronic Engineers (AIEE) in 1946. Another association was the Institute of Radio Engineers (IRE), which had a committee on electronic computers in 1951. The AIEE and IRE would eventually merge in 1963 to become the Institute of Electrical and Electronic Engineers (IEEE), which would include the IEEE Computer Society, whose headquarters are located in Washington, D.C. The IEEE Computer Society has numerous conferences and publishes thirteen peer-reviewed journals.
There are also computing and software associations in many other countries. These include the British Computer Society, the Computer Society of India, the Australian Computer Society, and many more. Computing and software have vast nets of social groups that share information via conferences, journals and, in today’s world, the web.
In 1947, two inventors, Thomas Goldsmith and Estie Ray Mann, filed a patent for using a cathode ray tube as a gaming device. This would later explode into a multibillion-dollar game industry within future decades.
The Cold War Begins
In 1949, the Soviet Union detonated its first atomic bomb, and this led to the concept of mutually
assured destruction, which meant that an atomic war on both sides would probably blast everyone
back to the Stone Age. The Soviet atomic bombs helped to keep the Cold War cool, since neither side wanted a full-scale atomic conflict.
The Soviet atomic bomb also had a major impact on computers and software. A direct response to the Soviet bomb was a new and massive air-defense system called SAGE, which stood for semiautomatic ground environment. SAGE will be discussed in the next chapter.
In 1949, Nationalist Chinese forces evacuated the mainland for Taiwan and the Chinese Communist government took over, which also led to tense relations with Western countries.
involved designing and building newer and more powerful computers of both digital and analog forms.
Postwar Computer Development
Two of the more technically important postwar computers were developed at Victoria University in Manchester, England. The first was the Manchester Mark I, also called the Small-Scale Experimental Machine (SSEM), which was operational in 1948.
The second and larger computer was the Manchester Automatic Digital Machine (MADM), which was operational in 1949. The British press called this machine an “electronic brain,” and this started a dispute between the engineering side of the university and the medical school. The dispute centered on whether computers could ever be creative.
The MADM led to 34 patents, some of which were later used in the IBM 701 and IBM 702 computers. The designers of the MADM were Frederic Williams and Tom Kilburn.
Kilburn and Williams filed a patent for a special kind of cathode ray tube called the Williams- Kilburn tube. It provided one of the first and fastest memory devices for storing digital data. Some of the early computers that used Williams-Kilburn tubes included the IBM 701 and IBM 702, the Univac Whirlwind, and the Ferranti Mark I.
When a dot is drawn on a cathode ray tube, it results in a positive charge, and the area surrounding the dot becomes negative. The charges spontaneously disperse, but they can be read and manipulated to store data. These tubes permitted random access, which was a major advance that opened up new kinds of computation.
Williams-Kilburn tubes were somewhat troublesome and not fully reliable. They were used during the late 1940s but were soon replaced by magnetic core memory devices in the early 1950s. Magnetic core memory was faster and more reliable.
However, when first introduced, magnetic cores had to be assembled by hand, using retrained garment workers who could deal with very small objects. In the early years of core memory, hundreds of garment workers in Europe and the United States were retrained to build computer core memories. As it happens, automated equipment for garment making occurred at about the same time, so otherwise the garment workers might have been laid off or unemployed.
Later in 1964, Dr. Robert Dennard of IBM’s Thomas J. Watson Research Center would receive U.S. patent 3,387,286 for the invention of DRAM, which would supplant older forms of computer memory. Several researchers working on core memory had filed patents, including Forrester and An Wang (who later founded Wang Laboratories). There were several years of patent disputes that were eventually resolved when IBM purchased all of the patents related to magnetic memory cores.
Another technically interesting postwar computer was the Electronic Delay Storage Automatic Calculator (EDSAC). This computer was built at the Mathematical Laboratory at Cambridge University in Great Britain. The designer was Maurice Wilkes and his colleagues. EDSAC was operational in May 1949 and was used to compute prime numbers, among other things.
As the name implies, the EDSAC used mercury delay lines for memory rather than cathode ray tubes. The mercury delay memory did not provide random access but was fairly reliable for sequential access.
An EDSAC programmer named David Wheeler is credited with having received the first Ph.D. in Computer Science in Great Britain. He is also credited with inventing the concept of subroutines. Subroutines would later be important in many programming languages, and they also led to the first creation of reusable code.
Some of the EDSAC library of 87 subroutines included floating point arithmetic, trigonometric functions, and exponentiation. Subroutines also allowed loops, and thus the do while loop was a feature of subroutines used on the EDSAC.
One of the most important postwar computers is the Whirlwind computer built by the Massachusetts Institute of Technology (MIT). The U.S. Navy approached MIT in 1944 about building a flight simulator for training bomber crews. The idea was to have a more realistic flight experience than was provided by the mechanical LINK trainer.
In 1947, Perry Crawford and Robert Everett completed the design of the Whirlwind (after prototype analog devices proved inadequate). The Whirlwind digital computer went operational on April 20, 1951.
The engineering team that built the Whirlwind worked for three years and included about 175 personnel, including seventy engineers and technicians. As can be seen, computers constructed from circuits that require hand soldering of connections are not easy to build. This is why the computer industry would be a small niche industry without the later development of integrated circuits in the next decade.
Among the novel features of the Whirlwind was the use of 16 math units operating in parallel, which made the Whirlwind sixteen times faster than computers using serial math.
The Whirlwind initially used mercury delay lines for memory. In a mercury delay line, a tube of liquid mercury had a microphone at one end and a transducer at the other end. Pulses were sent into the mercury and moved through it at the speed of sound until they were received at the other end. The signals were then amplified and sent back again, so the memory recirculated. The speed of sound varied with temperature, so the mercury delay lines did not operate at constant speeds, which caused problems. In addition to erratic performance, mercury is poisonous, so broken tubes were an occupational hazard of some significance.
Both mercury delay lines and cathode ray tubes were too slow and unreliable to be effective as computer memory devices. The project manager for Whirlwind was Jay Forrester. He had read about a new form of magnetic material and ordered samples. He experimented in his spare time at a workbench in the corner of the lab. After several months Forrester developed magnetic core memory. His first prototype consisted of thirty-two cores, each about 3/8 of an inch in diameter.
Forrester turned over the memory core project to a graduate student, and within two years, magnetic core memory was ready to go commercial to replace mercury delay lines and cathode ray tubes as the memory storage of choice for digital computers. Later, IBM developed the magnetic core concept as well as machines to speed up core memory construction.
The Whirlwind computer would be the basis of the SAGE air-defense system in the next decade and some of its technology would also find its way into SABRE, although improved by IBM. (SABRE stands for Semi-Automated Business Research Environment, which is a somewhat convoluted name perhaps selected merely to use the acronym “sabre.”)
An informal use of the Whirlwind computer was the development of a “bouncing ball” game in about 1949 by a researcher named Charley Adams. This was a precursor to later games such as Pong that would generate billions in revenue. It is interesting that computer games started to appear almost as soon as computers themselves.
Eckert and Mauchly formed the Eckert-Mauchly Computer Company (EMCC) in 1949 and it later became Univac. This was the world’s first pure digital computer company.
Northrup Grumman commissioned EMCC to build a computer for corporate use, and the result was the Binary Automatic Computer (BINAC). This computer had two separate processing units, each of which could hold 512 words. Mercury delay lines were used for memory storage.
Since the computer was commissioned for a client, it can be considered the world’s first commercial computer, but that is really stretching the definition of “commercial,” which normally implies multiple customers and multiple sales of the same product. The BINAC was definitely the first contract computer, but only one was built and there was only one customer.
As examples of how difficult and small computer programs were in this era, some of the BINAC test programs were each five to seven lines of code, and a “big” program during testing was 23 lines of code. The largest test program prior to delivery was 50 lines of code.
The BINAC was delivered to Northrup in September 1949, but it did not work properly after delivery. Northrup claimed the computer was not packed properly or was damaged in shipping. EMCC stated that the computer had probably not been assembled properly by Northrup, since EMCC personnel were not permitted onsite and assembly was performed by a graduate student without assistance from EMCC.