Where a calculator on the ENIAC is equipped with 18,000 vacuum tubes and weighs 30 tons, computers in the future may have only 1,000 vacuum tubes and weigh only 1 1/2 tons.

Popular Mechanics, 1949

Never trust a computer you can't lift.

Stan Mazor, 1970

If the automobile had followed the same development cycle as the computer, a Rolls-Royce would today cost $100, get one million miles to the gallon, and explode once a year, killing everyone inside.

Robert X. Cringely

Computers are such an integral part of our society that it is sometimes difficult to imagine life without them. However, computers as we know them are relatively new devices, with the first electronic computers dating back to the 1940's. Since that time, technology has advanced at an astounding rate, with the capacity and speed of computers approximately doubling every two years. Today, pocket calculators have many times the memory capacity and processing power of the mammoth computers of the 1950's and 1960's.

This chapter presents an overview of the history of computers and computer technology. The history of computers can be divided into generations, roughly defined by technological advances that led to improvements in design, efficiency, and ease of use.

Generation 0: Mechanical Computers (1642-1945)

Pascal's calculator

The first working calculating machine was invented in 1623 by German inventor Wilhelm Schickard (1592-1635), although details of its design were lost in a fire soon after its construction. In 1642, the French scientist Blaise Pascal (1623-1635) built a mechanical calculator when he was only 19. His machine used mechanical gears and was powered by hand. A person could enter numbers up to eight digits long using dials and then turn a crank to either add or subtract. Thirty years later, the German mathematician Gottfried Wilhelm von Leibniz (1646-1716) generalized the work of Pascal to build a mechanical calculator that could also multiply and divide. A variation of Leibniz's calculator, built by Thomas de Colmar (1785-1870) in 1820, was widely used throughout the 19th century.

Jacquard's loom

While mechanical calculators grew in popularity in the early 1800's, the first programmable machine built was not a calculator at all, but a loom. Around 1801, Frenchman Joseph-Marie Jacquard (1752-1834) invented a programmable loom that used removable punched cards to represent patterns. Before Jacquard's loom, producing tapestries was complex and tedious work. In order to produce a pattern, different colored threads (called wefts) had to be woven over and under the cross-threads (called warps) to produce the desired effect. Jacquard devised a way of encoding the patterns of the threads using metal cards with holes punched in them. When a card was fed through the machine, hooks passed through the holes to selectively raise warp threads, and so produce the desired over-and-under pattern. The result was that complex brocades could be encoded using the cards, and then reproduced exactly. Simply by changing the cards, the same loom could be used to automatically weave different patterns.

Babbage's Difference Engine

The idea of using punched cards for storing data was adopted in the mid 1800's by the English mathematician Charles Babbage (1791-1871). In 1821, he proposed the design of his Difference Engine, a steam-powered mechanical calculator for finding the solutions to polynomial equations. Although a fully functional model was never completed due to limitations in 19th century manufacturing technology, a prototype that punched output onto copper plates was built and used in computing data for naval navigation. Babbage's work with the Difference Engine led to his design in 1833 of a more powerful machine that included many of the features of modern computers. His Analytical Engine was to be a general-purpose, programmable computer that accepted input via punched cards and printed its output on paper. Similar to the design of modern computers, the Analytical Engine was to be made of integrated components, including a readable/writeable memory for storing data and programs (which Babbage called the store), and a control unit for fetching and executing instructions (which he called the mill). Although a working model of the Analytical Engine was never completed, its innovative and visionary design was popularized by the writings and patronage of Augusta Ada Byron, Countess of Lovelace (1815-1852).

Hollerith's machine

Punch cards resurfaced in the late 1800's in the form of Herman Hollerith's tabulating machine. Hollerith invented a machine for sorting and tabulating data for the 1890 U.S. Census. His machine utilized punch cards to represent census data, with specific holes on the cards representing specific information (such as male/female, age, home state, etc.) Recall that in Jacquard's loom, the holes in punch cards allowed hooks to selectively pass through and raise waft threads. In the case of Hollerith's tabulating machine, metal pegs passed through holes in the cards, making an electrical connection with a plate below that could be sensed by the machine. By specifying the desired pattern of holes, the machine could sort or count all of the cards corresponding to people with given characteristics (such as all men, aged 30-40, from Maryland). Using Hollerith's tabulating machine, the 1890 census was completed in six weeks (compared to the 7 years required for the 1880 census). Hollerith founded the Tabulating Machine Company in 1896 to market his machine. Eventually, under the leadership of Thomas J. Watson, Sr., Hollerith's company would become known as International Business Machines (IBM).

electromagnetic relay

It wasn't until the 1930's and the advent of electromagnetic relays that computer technology really started to develop. An electromagnetic relay is a mechanical switch that can be opened and closed by an electrical current that magnetizes it. The German engineer Konrad Zuse (1910-1995) is credited with building the first computer using relays in the late 1930's. However, his work was classified by the German government and eventually destroyed during World War II, and so did not influence other researchers. In the late 1930's, John Atanasoff (1903-1995) at Iowa State and George Stibitz (1904-1995) at Bell Labs independently designed and built automatic calculators using electromagnetic relays. In the early 1940's, Howard Aiken (1900-1973) at Harvard rediscovered the work of Babbage and began applying some of Babbage's ideas to modern technology. Aiken's Mark I, built in 1944, may be seen as an implementation of Babbage's Analytical Engine using relays.

In comparison with modern computers, the speed and computational power of these early computers may seem primitive. For example, the Mark I computer could store only 72 numbers in memory, although it could store an additional 60 constants through manual switches. The machine could perform 10 additions per second, but required up to 6 seconds to perform a multiplication and 12 seconds to perform a division. Still, it was estimated that complex calculations could be completed 100 times faster using the Mark I as opposed to existing technology at the time.

Generation 1: Vacuum Tubes (1945-1954)

vacuum tubes

While electromagnetic relays were certainly much faster than wheels and gears, they still required the opening and closing of mechanical switches. Thus, computing speed was limited by the inertia of moving parts. Relays also tended to be cumbersome and had a tendency to jam. A classic example of this is an incident involving the Harvard Mark II (1947), in which a computer failure was eventually traced to a moth that had become wedged between relay contacts. Grace Murray-Hopper (1906-1992), who was on Aiken's staff at the time, taped the moth into the computer logbook and facetiously noted the "First actual case of bug being found."

In the mid 1940's, electromagnetic relays began to be replaced by vacuum tubes, resulting in machines whose only moving parts were electrons. Invented by Lee de Forest (1873-1961) in 1906, a vacuum tube is a small glass tube from which all or most of the gas has been removed, permitting electrons to move with minimal interference from gas molecules. Since vacuum tubes have no moving parts (only the electrons move), they permitted the switching of electrical signals at speeds far exceeding those of any mechanical device. In fact, vacuum tubes were up to 1,000 times faster than electromagnetic relays. They could also be used to construct fast storage devices.

COLOSSUS

The stimulus for the electronic (vacuum tube) computer was World War II. The first electronic computer was COLOSSUS, built by British government to decode encrypted Nazi communications. COLOSSUS was designed in part by the British mathematician Alan Turing (1912-1954), who is considered one of the founding fathers of computer science due to his seminal work in the theory of computability and artificial intelligence. While operational in 1943, COLOSSUS's influence over other researchers at the time was limited as its existence remained classified for over 30 years. It contained more than 2,300 vacuum tubes, and had a unique design due to its intended purpose as a code-breaker. It contained 5 different processing units, each of which could read in and process 5,000 characters of code per second. Using COLOSSUS, British Intelligence was able to decode Nazi military communications for extended periods, providing invaluable support to Allied operations during the war.

ENIAC

At roughly the same time in the United States, physics professor John Mauchly (1907-1980) and his graduate student J. Presper Eckert (1919-1995) designed ENIAC, which was an acronym for Electronic Numerical Integrator And Computer. The ENIAC was designed to compute ballistics tables for the U.S. Army, but it was not completed until 1946. It consisted of 18,000 vacuum tubes and 1500 relays, weighed 30 tons, and consumed 140 kilowatts of power. In contrast to the Mark I, the ENIAC could only store 20 numbers in memory, although more than 100 constants could be entered using switches. Still, the ENIAC could perform more complex calculations than the Mark I and operated up to 500 times faster (5,000 additions per second). The ENIAC was programmable in that it could be reconfigured to perform different computations. However, reprogramming the machine required manually setting up to 6,000 multiposition switches and reconnecting cables. In a sense, it was not so much reprogrammed as it was rewired each time the application changed.

John von Neumann

One of the scientists involved in the ENIAC project was John von Neumann (1903-1957), who along with Turing is considered a founding father of the field of computer science. Von Neumann recognized that programming via switches and cables was tedious and error prone. As an alternative, he designed a computer architecture in which the program could be stored in memory along with the data. Although this idea was initially proposed by Babbage with his Analytical Engine, von Neumann is credited with its formalization using modern designs. Von Neumann also recognized the advantages of using binary (base 2) representation in memory as opposed to decimal (base 10), which had been used previously. His basic design was used in EDVAC (Eckert and Mauchly at UPenn, 1952), IAS (von Neumann at Princeton, 1952), and subsequent machines. The von Neumann architecture is still the basis for nearly all modern computers.

As a result of von Neumann's "stored program" architecture, the process of programming a computer became equally if not more important than designing a computer. Before von Neumann, computers were not so much programmed as they were hard-wired to a particular task. With the von Neumann architecture, a program could be read in (via cards or tapes) and stored in the memory of the computer. Reprogramming the computer did not require any reconfiguring, it simply involved reading in and storing a new program for execution. At first, programs were written in machine language, sequences of 0's and 1's that corresponded to instructions directly executable by the hardware. While this was an improvement over rewiring, it still required the programmer to write and manipulate pages of binary numbers -- a formidable and error-prone task. In the early 1950s, assembly languages were invented which simplified the programmer's task somewhat by substituting simple mnemonic names for binary numbers.

The early 1950's also marked the birth of the commercial computer industry. Eckert and Mauchly left the University of Pennsylvania to form their own company. In 1951, The Eckert-Mauchly Computer Corporation (later a part of Remington-Rand, then Sperry-Rand) began selling the UNIVAC I computer. The first UNIVAC I was purchased by the U.S. Census Bureau, and a subsequent UNIVAC I captured the public imagination when it was used by CBS to predict the 1952 presidential election. Several other companies would soon follow Eckert-Mauchly with commercial computers. It is interesting to note that at this time, International Business Machines (IBM) was a small company producing card punches and mechanical card sorting machines. It entered the computer industry late in 1953, and did not really begin its rise to prominence until the early 1960's.

Generation 2: Transistors (1954-1963)

RCA transistor ad from Fortune 1953/3

Vacuum tubes, in addition to being relatively large (several inches long), dissipated an enormous amount of heat. Thus, they required lots of space for cooling and tended to burn out frequently. The next improvement in computer technology was the replacement of vacuum tubes with transistors. Invented by John Bardeen, Walter Brattain, and William Shockley in 1948, a transistor is a piece of silicon whose conductivity can be turned on and off using an electric current. Transistors were much smaller, cheaper, more reliable, and more energy-efficient than vacuum tubes. As such, they allowed for the design of smaller and faster machines at drastically lower cost.

Transistors may be the most important technological development of the 20th century. They allowed for the development of small and affordable electronic devices: radios, televisions, phones, computers, etc., as well as the information-based economy that developed along with them. The potential impact of transistors was recognized almost immediately, earning Bardeen, Brattain, and Shockley the 1956 Nobel Prize in physics. The first transistorized computers were supercomputers commissioned by the Atomic Energy Commission for nuclear research in 1956: Sperry-Rand's LARC and IBM's STRETCH. Companies such as IBM, Sperry-Rand, and Digital Equipment Corporation (DEC) began marketing transistor-based computers in the early 1960's, with the DEC PDP-1 and IBM 1400 series being especially popular.

Rear Admiral Grace Murray-Hopper

As transistors drastically reduced the cost of computers, even more attention was placed on programming. If computers were to be used by more than just the engineering experts, programming would have to become simpler. In 1957, John Backus (1924-) and his group at IBM introduced the first high-level programming language, FORTRAN (FORmula TRANslator). This language allowed the programmer to work at a higher level of abstraction, programming via mathematical formulas as opposed to assembly level instructions. Programming in a high-level language like FORTRAN greatly simplified the task of programming, although IBM's original claims that FORTRAN would "eliminate coding errors and the debugging process" were a bit optimistic. Other high-level languages were developed in this period, including LISP (John McCarthy at MIT, 1959), BASIC (John Kemeny at Dartmouth, 1959), and COBOL (Grace Murray-Hopper at the Department of Defense, 1960).

The 1960's saw the rise of the computer industry, with IBM becoming a major force. The success of IBM is not so much attributed to superior technology, but more to its shrewd business sense and effective long-range planning. While smaller companies like DEC, CDC, and Burroughs may have had more advanced machines, IBM won market share through aggressive marketing, reliable customer support, and a strategy of designing an entire family of compatible, interchangeable machines.

Generation 3: Integrated Circuits (1963-1973)

integrated circuits

Transistors were far superior to the vacuum tubes they replaced in many ways: they were smaller, cheaper to mass produce, and more energy-efficient. However, early transistors still generated a great deal of heat, which could damage other components inside the computer. In 1958, Jack Kilby (1923-) at Texas Instruments developed a technique for manufacturing transistors as layers of conductive material on a silicon disc. This new form of transistor could be made much smaller than existing, free-standing transistors, and also generated less heat. Thus, tens or even hundreds of transistors could be layered onto the same disc and connected with conductive layers to form simple circuits. Such a disc, packaged in metal or plastic with external pins for connecting with other components, is known as an Integrated Circuit (IC) or IC chip. Packaging transistors and related circuitry on an IC that could be mass-produced made it possible to build computers that were smaller, faster, and cheaper. A complete computer system could be constructed by mounting a set of IC chips (and other support devices) on circuit boards and soldering them together.

Intel 4004

As manufacturing technology improved, the number of transistors that could be fit on a single chip increased. In 1965, Gordon Moore (1929-) noticed that the number of transistors that could fit on a chip roughly doubled every 12 to 18 months. This trend became known as Moore's Law, and has continued to be an accurate predictor of technological advances. By the 1970's, the Large Scale Integration (LSI) of thousands of transistors on a single IC chip became possible. In 1971, the Intel Corporation made the logical step of combining all of the control circuitry for a calculator into a single chip called a microprocessor. The Intel 4004 microprocessor contained more than 2,300 transistors. Its descendant, the 8080, which was released in 1974, contained 6,000 transistors and was the first general-purpose microprocessor, meaning that it could be programmed to serve various functions. The Intel 8080 and its successors, the 8086 and 8088 chips, served as central processing units for numerous personal computers in the 1970's, including the IBM PC. Other semiconductor vendors such as Texas Instruments, National Semiconductors, Fairchild Semiconductors, and Motorola also produced microprocessors in the 1970's.

As was the case with previous technological advances, the development of integrated circuits led to faster and cheaper computers. Whereas computers in the early 1960's were affordable only for large corporations, IC technology rapidly brought down costs to the level where small businesses could afford them. As such, more and more people needed to be able to interact with computers. A key to making computers easier for non-technical users was the development of operating systems, master control programs that oversee the operation of the computer and manage peripheral devices. An operating system acts as an interface between the user and the physical components of the computer, and through a technique known as time-sharing, can even allow multiple users to share the same computer simultaneously. In addition to operating systems, specialized programming languages were developed to fill the needs of the broader base of users. In 1971, Niklaus Wirth (1934-) developed Pascal, a simple language designed primarily for teaching but which has dedicated users to this day. In 1972, Dennis Ritchie (1941-) developed C, a systems language used in the development of the UNIX operating system. Partly because of the success of UNIX, C and its descendant C++ have become the industry standards for systems programming and software engineering.

Generation 4: VLSI (1973-1985)

Intel Family of IC Chips

Year	Processor	# Transistors
2000	Pentium 4	42,000,000
1999	Pentium III	9,500,000
1997	Pentium II	7,500,000
1993	Pentium	3,100,000
1989	80486	1,200,000
1985	80386	275,000
1982	80286	134,000
1978	8088	29,000
1974	8080	6,000
1972	8008	3,500
1971	4004	2,300

While the previous generations of computers were defined by new technologies, the jump from generation 3 to generation 4 is largely based upon scale. By the mid 1970's, advances in manufacturing technology led to the Very Large Scale Integration (VLSI) of hundreds of thousands and eventually millions of transistors on an IC chip. As an example of this, consider the table to the left that shows the number of transistors on chips in the Intel family. It is interesting to note that the first microprocessor, the 4004, had approximately the same number of switching devices (transistors) as the COLOSSUS, which utilized vacuum tubes in 1943. Throughout the 1970's and 1980's, successive generations of IC chips added more and more transistors, providing more complex functionality in the same amount of space. It is astounding to note that the Pentium 4, which was released in 2000, has more than 42 million transistors, with individual transistors as small as 0.18 microns (0.00000000018 meters).

As the mass production of microprocessor chips became possible in the 1970's with VLSI, the cost of computers dropped to the point where a single individual could afford one. The first personal computer (PC), the MITS Altair 8800, was marketed in 1975 for under $500. In reality, the Altair was a computer kit that included all of the necessary electronic components, including the 8080 microprocessor that served as the central processing unit for the machine. The customer was responsible for wiring and soldering all of these components together to assemble the computer. Once assembled, the Altair had no keyboard, no monitor, and no permanent storage. The user entered instructions directly by flipping switches on the console, and viewed output as blinking lights. Despite these limitations, demand for the Altair was overwhelming.

Jobs & Wozniak

While the company that sold the Altair, MITS, would fold within a few years, other small computer companies successfully charted the PC market in the late 1970's. In 1976, Steven Jobs (1955-) and Stephen Wozniak (1950-) got their start selling computer kits similar to the Altair, which they called the Apple. In 1977, they founded Apple Computer and began marketing the Apple II, the first pre-assembled personal computer with a keyboard, color monitor, sound, and graphics. By 1980, the annual sales of their Apple II personal computer had reached almost $120 million. Other companies, such as Tandy, Amiga, and Commodore soon entered the market with personal computers of their own. IBM, which had been slow to enter the personal computer market, introduced the IBM PC in 1980, becoming an immediate force. Apple countered with the Macintosh in 1984, which introduced the now familiar graphical user interface of windows, icons, and a mouse.

Bill Gates

As computers became the tools of more and more individuals, the software industry grew and adapted. Bill Gates (1955-) is credited with writing the first commercial software for personal computers, a BASIC interpreter for the Altair. In 1975, while a freshman at Harvard, he cofounded Microsoft along with Paul Allen (1955-) and helped to build that company into the software giant that it is today. Much of Microsoft's initial success was due to its marketing of the MS-DOS operating system for PCs, and applications programs such as word processors and spreadsheets. By the mid 1990's, Microsoft Windows had become the dominant operating system for desktop computers, and Bill Gates, in his early 40's, had become the richest person in the world.

The proliferation of programming languages continued in the 1980's and beyond. In 1980, Alan Kay (1940-) developed Smalltalk, the first object-oriented language. Ada was developed for the Department of Defense in 1980, to be used for all government contracts. In 1985, Bjarne Stroustrup (1950-) developed C++, an object-oriented extension of C. C++ and its offshoot Java (developed in 1995 at Sun Microsystems) have become the dominant languages in software engineering today.

Generation 5: Parallel Processing & Networking (1985-????)

Deep Blue defeats Kasparov

While generations 0 through 4 are reasonably well-defined, the scope of the fifth generation of computer technology is still an issue of debate. What is clear is that computing in the late 1980's and beyond has been defined by advances in parallel processing and networking. Parallel processing refers to the integration of multiple (sometimes hundreds or thousands) processors in a computer. By sharing the computational load across multiple processors, a parallel computer is able to execute programs in only a fraction of the time required by a single processor computer. A striking example of this is IBM's Deep Blue, a parallel processing computer specifically designed for playing chess. Using 256 processors, Deep Blue is able to evaluate millions of potential moves in a second and thus select the most promising move. In 1997, Deep Blue became the first computer to defeat the reigning world's chess champion in tournament play.

Up until the 1990's, most computers were stand-alone devices. Small-scale networks of computers were common in large businesses, but communication between networks was rare. The idea of a large-scale network, connecting computers from remote sites, traces back to the 1960's. The ARPAnet, connecting four computers in California and Utah, was founded in 1969. However, its use was mainly limited to government and academic researchers. As such, the ARPAnet, or Internet as it would later be called, grew at a slow but steady pace. That changed in the 1990's with the development of the World Wide Web. The Web, a multimedia environment in which documents can be seamlessly linked together, was introduced in 1989 by Tim Berners-Lee, but first became popular in the mid 1990's with the development of graphical Web browsers. As of 2002, it is estimated that there are more than 160 million computers connected to the Internet, of which more than 33 million are Web servers storing up to 5 billion individual Web pages. For more details on the development of the Internet and Web, refer to Chapter 3.

(Optional) Supplementary Readings

• Jones Multimedia Encyclopedia: www.digitalcentury.com
• Computer Museum of America: www.computerhistory.org
• Charles Babbage Institute: www.cbi.umn.edu/tc.html
• Bletchley Park: www.bletchleypark.org
• ENIAC Museum: www.seas.upenn.edu /~mu seum/ hist-overview.html
• PBS History of the Transistor: www.pbs.org/transistor/
• Internet Software Consortium: www.isc.org/ds/
• Netcraft Web Server Survey: www.netcraft.com

Review Questions

Answers to be submitted to your instructor via email by Thursday at 9 a.m.

1. Jacquard's loom, while unrelated to computing, was influential in the development of modern computing devices. What design features of that machine are relevant to the design of modern computers?
2. What role did Herman Hollerith play in the history of computer technology?
3. What advantages did vacuum tubes have over electromagnetic relays?
4. One advantage of the von Neumann architecture over previous computer architectures was that it was a stored-program machine. Describe how such a computer could be programmed to perform different computations without physically rewiring it. co
5. What is a transistor and how did the introduction of transistors lead to faster and cheaper computers?
6. What is Moore's Law?
7. What does the acronym VLSI stand for?
8. What was the first personal computer and when was it first marketed?
9. How much did you know about the history of computers before reading this chapter? Did you find it informative?
10. What did you find most interesting in reading this chapter?
11. What did you find most confusing in reading this chapter?