Computers-Small computers: Past, present, and future.

In this presentation I will try to review some of the progress in the computing field over the last 10 or 12 years, especially in the minicomputer and small computer area-that is, desk-top and hand-held machines. Then, on the basis of this progress and what we already know about further developments in microelectronics, I would like to offer some predictions of things to come in the decade ahead. Hewlett-Packard backed into the minicomputer business a little over 12 years ago. I wish I could say that we foresaw the great future that the minicomputer was going to have, but we did not. We built one as a controller in order to make automatic measuring systems. That controller and its descendants now have become half of our business. So this just shows you what a rapidly growing field it is. Most of the rapid progress that has been made in computers over the last decade has been due to the great reductions in size and cost of the hardware. This is well illustrated by what has happened in memories. Consider the trends there. Back in 1968, 22-mil magnetic cores were used for memory. In 1971 the core size was reduced to 18 mils. In 1974, "large-scale integrated circuit" memories appeared with 4000 bits of storage; these were followed in 1977 with 16,000-bit chips. In 12 years, size has decreased 60-fold, power requirement has decreased 25fold, and the cost has decreased about 55-fold. Over the same period the logic used in computers changed from current transfer logic (CTL) to transistor-transistor logic (TTL) and to Shocky-clamped transistor-transistor logic (CTTL) to large-scale integrated (LSI) circuits made up of complementary MOS transistors or silicon-on-sapphire transistors. There has been a corresponding decrease in price, size, and weight of these computers over this period along with an increase in performance. Together with reduced memory costs, these changes have allowed the price and power consumption of a 64-kilobyte machine to decrease 12-fold, its volume to decrease 10-fold, and its weight to decrease almost 10-fold. We can look at this another way. If in 1968 you had $50,000 to spend, you could have become the proud owner of a 64-kilobyte machine and nothing else, not even input or output devices. Today, $50,000 is worth about half as much for most things we buy, but it will buy you a minicomputer with 256 kilobytes of memory and more than 10 times the performance, plus a 50-megabyte disk and a terminal with graphics capability. There has been a great improvement and a great increase in what the system will do for you. In addition to more memory and to more software and hardware features, there has been a steady increase in the operating system capabilities and the ease with which you can interface with it. In some ways the rapid evolution of the desk-top computer or calculator into a computer is even more dramatic than the evolution of the minicomputer. In 1968 we introduced the first programmable desk-top calculator with transcendental functions available at a keystroke. It had 4000 bytes of read-only memory in the form of a 14-layer printed circuit board, but it had only 276 bytes of random access core memory. This was doubled 18 months later in the 9100B. Although these machines filled a real need, their data space was too small for most applications. The advent of the 4K random access memories and ROMs as large-scale integrated chips made possible a new generation of desk-top machines with high-level programming languages as well as greatly increased memory. The 9810-20 and -30 had memories ranging from 1 to 30 kilobytes and so were entering the minicomputer range. A third generation of machines offered about a 10-fold increase in speed as well as graphics capability. These computers had memories as great as 450 kilobytes-like a large computer of the 1960s. That represents a 1600-fold increase over the 9100 A produced 10 years earlier. In summary, over the 12-year period we have seen an advance from 276 bytes to 32,000 bytes and from a 3-inch cathode ray tube that only displayed numbers to a 5-inch tube that displays alphanumerics and graphics. With the early machines you could buy a printer at extra cost. Current ones have it self-contained. The programming language has grown from reversed Polish notation to BASIC. The increase in capability has been accompanied by a decrease in price from $4900 to $3600. The same story of increasing power holds for the hand-held calculator. From the original HP-35 to the programmable 65 and 67 to the latest 41C, we see the following trends: a 15-fold increase in internal programs and algorithms as represented by the amount of ROM (can be 50-fold at an added cost); an 80fold increase in data storage as indicated by the amount of random access memory; and a 25% reduction in price. So much for the 1970s. I have not been trying to give a sales pitch. As a matter of fact, I have been doing just the opposite. Because the only conclusion you could logically draw from what I've said is: "No matter what year it is, don't buy a computer or calculator; wait until next year." This is Oliver's second law. And I am afraid that it holds for the 1980s. The enabling technology behind the dramatic increase in computing power per dollar that has taken place and that is continuing to take place is that of the large-scale circuit integration. Today's circuits that give us 16,000 bits per memory chip, for example, are made with 4-,um lines and spaces. The technology is already at hand to reduce this to 2-,gm lines and spaces, which would give us 64,000-bit memories at the same price we now pay for 16,000-bit memories. Further size reductions are in the offing. Line widths of 1 um are currently being drawn with electron beam mask-making machines. Certainly, line widths finer than 1 Am are possible, and so the question comes up immediately where is all this going to end? I believe it will end when the scale of integration is limited not by the technology of lithography, for that will continue to be improved, but rather by the physics of the device itself. When devices get too small, the statistical fluctuations in doping become a problem. The allowable working voltages become too small compared with kT/q. Depletion layer punch-through occurs easily. Electrons tunnel through gates and through gate