Microchip

Although semiconductors and microchips are essential components of modern computers, many people do not realize that computing machinery does not really need to be constructed with components that are normally associated with electronic equipment. In fact, some of the earliest computers were purely mechanical machines—they did not rely on electrical technology at all. For example, Charles Babbage's Analytical Engine, designed in 1834 at a time when the use of electricity was in its infancy, was a purely mechanical machine. Had Babbage actually been able to build it, his Analytical Engine would have been a bona fide computing machine.

Similarly, many of the early computers and calculators were mostly mechanical, using carefully constructed linkages, levers, and cogs. It is important to note that the technology used to implement computers does not define them. Instead, machines are termed computers if they are programmable—regardless of the form the programming takes. Therefore, once mechanical computers and calculators had proven themselves somewhat cumbersome and inefficient, designers looked toward the then newly emerging electro-technologies as a means for implementing computers and calculators.

Around the mid-twentieth century, the analog computer was becoming an increasingly popular tool for solving differential equations. Valve and triode devices used in analog amplification equipment were being mass-produced for the radio and wireless sets that were consumer items of the day. They were also suitable building blocks for the implementation of analog computers. Yet, while analog computers were predecessors of modern digital computers, they did not bear much resemblance to current digital computers.

To explain the development of these technologies, it is helpful to analyze their development history. Scientists and mathematicians have known since the eighteenth century that differential and integral calculus can be used to model problems in the physical sciences. Also, while solutions to differential equations can be developed manually, this process tends to be tedious. Analog computers offered a way of automating the process of generating solutions to differential and integral equations. Building blocks made from valves and triodes were constructed to perform specific operations that are common in the solution of differential and integral equations. Blocks that could complete arithmetic operations—such as addition, subtraction, multiplication, and division—could be assembled, along with others that affected operations like integration, differentiation, and other forms of filtering. These building blocks could be assembled and connected using temporary wiring connections—this was actually the programming of these computers.

In the beginning, programming an analog computer was a rather labor-intensive activity and the computers themselves would consume a relatively large amount of electrical power. But their speed of computation was phenomenal compared to mechanical computers. The valves and triodes that made these machines possible are still occasionally found in esoteric modern audio amplifier equipment, but have been largely consigned to history. The cause of this was the invention of the semiconductor transistor device in 1948.

Solid-state physicist William B. Shockley (1910–1989) described the operation of the semiconductor transistor in 1950, and its development foreshadowed a revolution in electronics. The fundamental physical difference between a conductor of electricity and an insulator is that conductors permit free flow of electrons, and insulators do not. In other words, if someone takes a piece of conducting material, like a metal, and drops a packet of electrons onto it at one point, they will almost instantaneously redistribute themselves throughout the volume of the metal sample. They will tend to spread out so that their distribution is uniform. A piece of insulating material, like polyvinyl chloride (PVC plastic), would tend to resist the redistribution of a packet of electrons. The PVC would try to prevent the localized collection of electrons from redistributing themselves—instead they would be contained in the one area making that region negatively charged.

Shockley and his contemporaries discovered that there was a certain class of materials that could sometimes be seen as acting like conductors, but with a certain amount of manipulation, the same material could be made to act as an insulator. This property made them somewhat special—they could conduct or insulate under control, making them ideal as switching devices. These materials became known as semiconductors, as a result of their unusual position logically between conductors and insulators. Silicon and germanium were identified early as semiconductor materials.

The production process for the creation of a semiconductor is a complex multistage activity, but essentially involves minuscule semiconductor elements being impregnated with charged particles (known as doping) so as to influence their behavior in useful ways. They are bonded to conductors and encased in plastic or ceramic containers ready for use. Since this time, the word "silicon" in the context of electronics has been used synonymously with terms such as "silicon chip," "chip," and "microchip."

Silicon, germanium, and other semiconductor materials derived from metal oxides have been used ever since, along with metals such as gold, aluminium, and copper to produce semiconductor integrated circuit devices of extraordinary complexity and performance. Their successful miniaturization has meant that a great deal of functionality can be synthesized on a relatively small device. Additionally, these devices consume much less electrical power and operate at vastly greater speeds than the older valve and triode devices. Extra benefits have resulted from the perfection of the manufacturing processes as well, which has in turn lead to these devices becoming inexpensive to purchase and reliable in operation.

An entire industry of massive proportions has been supported by these developments, with its genesis in an area near San Francisco, California, which has since become known as Silicon Valley. Subsequently, other regions in Europe and Asia—notably Japan and South Korea—have also established credibility in the mass-production of semiconductors.

For some time theorists and visionaries have proposed the idea that semiconductors might eventually be replaced in computers by devices that have the capacity to implement computer circuitry by using optics or quantum physical concepts, but these are yet to be proven beyond the research laboratory. Any replacement technology will need to possess very impressive credentials indeed if it is to be as operationally effective, as economical and efficient as devices implemented from semiconductors.