In 1955, an early high-speed commercial computer weighed 3 tonnes, consumed 50 kilowatts of power, and cost $400,000. But it could perform 50 multiplications per second, a feat unmatchable by either a human or the latest adding machine. In 1977, a handheld calculator weighed under about 0.5 kg, consumed less than half a watt of power, could perform 250 multiplications per second, and cost $600. Today, you can buy palm-sized organizers for $250 that link to computers, transmit data, and store thousands of addresses, appointments, memos, lists, and e-mails. The keys to this stunning revolution in personal power are the transistor and the integrated circuit -- the centrepieces of the modern electronics systems that swept the world in the last half of the 20th century. Brilliant engineering and innovation lie behind these unseen elements that operate wireless communications, satellite broadcasts, air traffic control systems, microwave ovens, video cameras, touch-tone phones, computers, and many other products that have improved the quality, safety, and convenience of modern life.
The ancestor of these miniature electronic devices is the vacuum tube. Sealed inside a glass tube, a stream of electrons carried a current through a vacuum between electrodes. Vacuum tubes were crucial to the development of radio, television, and sound recording, and an essential component in early telephone equipment and computers. They were also fragile, bulky, and produced a considerable amount of waste heat. The first commercial computer, ENIAC, incorporated 18,000 vacuum tubes, weighed 30 tonnes, filled several large rooms, and consumed enough power to light 10 homes. Its cathode ray tubes required large amounts of heat in order to boil out electrons, needed time to warm up, and often burned out. A universal search to find a more compact and reliable device dominated engineering after World War II, and these efforts laid the groundwork for what followed.
What followed was the transistor, invented in 1947 by John Bardeen, Walter H. Brattain, and William B. Shockley, engineers and scientists at Bell Telephone Laboratories. The transistor's shape and size were strikingly different from the huge arrangement of the bulky vacuum tubes. A small metal cylinder about half an inch long contained two fine wires that ran down to a pinhead of solid semiconductive material soldered to a metal base. The current to the crystal on one wire controlled a larger current between the crystal and the second wire. It contained no vacuum, grid, plate, or glass envelope to keep the air away. It produced instantaneous action, with no need to warm up. The New York Times reported its debut with only slight interest. Little did anyone realize this tiny device would launch the "smaller, faster, more powerful" digital age.
By the early 1950s, the transistor had captured the world's imagination, first in the transistorised radio - the fastest selling retail object of the time. Early applications included telephone oscillators, hearing aids, automatic telephone routing devices, and other audio and communications devices. Computers were not yet considered a key application. IBM could not find anyone interested in selling transistors that were tailored to computers, so they contracted with Texas Instruments to develop transistors specifically designed for digital applications.
The transistor held great promise, but it would take several years before engineers worked out designs and manufacturing techniques. Semiconductor material was costly and required complex contacting methods. To help speed progress, Bell Labs licensed its transistor patent rights freely to other companies. The rapid advance made in transistor manufacturing from 1952 to 1960 has been attributed to this open sharing of technology, and the subsequent developments made at various industry and university laboratories and presented at open symposiums.
These advances undoubtedly led to Jack Kilby's invention of the integrated circuit (IC) at Texas Instruments in 1958. At the time, miniaturisation was driven by the Soviet's success with the Sputnik program, and was a major objective of government-funded electronics research programs. Kilby came up with the idea of organising numerous transistors and other electronic components on a silicon wafer, complete with wiring. It would take much ingenuity and effort, and the adaptation of techniques learned from earlier transistor fabrication, such as crystal growing.
Early transistors were made by hand. Eventually, sophisticated production techniques took hold. The real breakthrough in production was the use of oxide on silicon wafers, which allowed selective doping by diffusion of impurities through openings in the oxide. This process allowed contacts to the silicon to be made through other holes in the oxide. Photographic techniques were used to pattern the openings in the silicon oxide. With these techniques, hundreds of chips could be cut from a single slice of silicon and multiple transistors could be placed on each chip. Early chips were about three-quarters of a millimetre on a side. Today chips are several centimetres on a side, and can accommodate millions of transistors.
In the early 1950s, a transistor cost between $10 and $90 to make. As the semiconductor technology improved, the transistor became faster, cheaper, and more reliable. Now the transistors on a microchip cost less than a hundred-thousandth of a cent - so they are virtually free.
At first, integrated circuits were produced by the hundreds. Then engineers developed ways to add other components - resistors and capacitors -- to produce a microchip. As the technology developed, more and more components could be crammed into smaller and smaller dimensions. Gordon Moore, chairman of Intel, recognized a trend: the number of transistors per unit was doubling every year, and later, every 18 months. This insight became known as Moore's Law, one of the driving principles of the semiconductor industry, and Moore's vision helped Intel become one of the world's major corporations.
Part of the magic of electronics is adding millions of transistors to a tiny silicon chip. The rest of the magic is performed by engineers who determine their use through the development of microprocessors - the control centre embedded in refrigerators, automobiles, airplanes, computers, and thousands of other products.
Microchips took the transistor to an exciting new level. One microchip can operate an automobile's electrical system or launch an air force. It made thousands of new products possible, from heart pacemakers and hearing aids to efficient aircraft. Medical instruments, automobiles, cellular phones, CD players, and watches all operate because of microchips.
Until microprocessors appeared on the scene, computers were essentially discrete pieces of equipment used primarily for data processing and scientific calculations. From microprocessors engineers developed microcomputers -- systems about the size of a lunch box or smaller, but with enough computing power to perform many kinds of business, industrial, and scientific tasks.
The race continues to add more and more information on a microchip. By 2010, advanced microprocessors are expected to contain more than 800 million transistors. Where will microchips appear next? They might appear on the front of refrigerators to monitor food supplies and send grocery lists to the store, automatically charging credit cards or bank accounts. Or they could be implanted in children to prevent kidnapping, or inside the human brain to cure blindness or other medical conditions. The technology is limitless. Only imagination will govern its potential.
***
***
***