Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 169
C Reprint of Gordon E. Moore’s “Cramming More Components onto Integrated Circuits” NOTE: Reprinted from Gordon Moore, 1965, Cramming more components onto integrated circuits, Electronics 38(8) with permission from Intel Corporation. 169
OCR for page 170
170 THE FUTURE OF COMPUTING PERFORMANCE Cramming More Components onto Integrated Circuits GORDON E. MOORE, LIFE FELLOW, IEEE Each approach evolved rapidly and converged so that With unit cost falling as the number of components per circuit rises, by 1975 economics may dictate squeezing as many as 65 000 each borrowed techniques from another. Many researchers components on a single silicon chip. believe the way of the future to be a combination of the The future of integrated electronics is the future of various approaches. electronics itself. The advantages of integration will bring The advocates of semiconductor integrated circuitry are about a proliferation of electronics, pushing this science already using the improved characteristics of thin-lm into many new areas. resistors by applying such lms directly to an active semi- Integrated circuits will lead to such wonders as home conductor substrate. Those advocating a technology based computers—or at least terminals connected to a central upon lms are developing sophisticated techniques for the computer—automatic controls for automobiles, and per- attachment of active semiconductor devices to the passive sonal portable communications equipment. The electronic lm arrays. wristwatch needs only a display to be feasible today. Both approaches have worked well and are being used But the biggest potential lies in the production of large in equipment today. systems. In telephone communications, integrated circuits in digital lters will separate channels on multiplex equip- II. THE ESTABLISHMENT ment. Integrated circuits will also switch telephone circuits Integrated electronics is established today. Its techniques and perform data processing. are almost mandatory for new military systems, since the Computers will be more powerful, and will be organized reliability, size, and weight required by some of them is in completely different ways. For example, memories built achievable only with integration. Such programs as Apollo, of integrated electronics may be distributed throughout for manned moon ight, have demonstrated the reliability the machine instead of being concentrated in a central of integrated electronics by showing that complete circuit unit. In addition, the improved reliability made possible functions are as free from failure as the best individual by integrated circuits will allow the construction of larger transistors. processing units. Machines similar to those in existence Most companies in the commercial computer eld have today will be built at lower costs and with faster turn- machines in design or in early production employing inte- around. grated electronics. These machines cost less and perform better than those which use “conventional” electronics. Instruments of various sorts, especially the rapidly in- I. PRESENT AND FUTURE creasing numbers employing digital techniques, are starting By integrated electronics, I mean all the various tech- to use integration because it cuts costs of both manufacture nologies which are referred to as microelectronics today and design. as well as any additional ones that result in electronics The use of linear integrated circuitry is still restricted functions supplied to the user as irreducible units. These primarily to the military. Such integrated functions are ex- technologies were rst investigated in the late 1950’s. The pensive and not available in the variety required to satisfy a object was to miniaturize electronics equipment to include major fraction of linear electronics. But the rst applications increasingly complex electronic functions in limited space are beginning to appear in commercial electronics, partic- with minimum weight. Several approaches evolved, includ- ularly in equipment which needs low-frequency ampliers ing microassembly techniques for individual components, of small size. thin-lm structures, and semiconductor integrated circuits. III. RELIABILITY COUNTS Reprinted from Gordon E. Moore, “Cramming More Components onto In almost every case, integrated electronics has demon- Integrated Circuits,” Electronics, pp. 114–117, April 19, 1965. strated high reliability. Even at the present level of pro- Publisher Item Identier S 0018-9219(98)00753-1. 82 PROCEEDINGS OF THE IEEE, VOL. 86, NO. 1, JANUARY 1998
OCR for page 171
171 APPENDIX C duction—low compared to that of discrete components—it offers reduced systems cost, and in many systems improved performance has been realized. Integrated electronics will make electronic techniques more generally available throughout all of society, perform- ing many functions that presently are done inadequately by other techniques or not done at all. The principal advantages will be lower costs and greatly simplied design—payoffs from a ready supply of low-cost functional packages. For most applications, semiconductor integrated circuits will predominate. Semiconductor devices are the only rea- sonable candidates presently in existence for the active elements of integrated circuits. Passive semiconductor el- ements look attractive too, because of their potential for low cost and high reliability, but they can be used only if precision is not a prime requisite. Silicon is likely to remain the basic material, although Fig. 1. others will be of use in specic applications. For example, gallium arsenide will be important in integrated microwave functions. But silicon will predominate at lower frequencies because of the technology which has already evolved V. TWO-MIL SQUARES around it and its oxide, and because it is an abundant and With the dimensional tolerances already being employed relatively inexpensive starting material. in integrated circuits, isolated high-performance transistors can be built on centers two-thousandths of an inch apart. Such a two-mil square can also contain several kilohms IV. COSTS AND CURVES of resistance or a few diodes. This allows at least 500 Reduced cost is one of the big attractions of integrated components per linear inch or a quarter million per square electronics, and the cost advantage continues to increase inch. Thus, 65 000 components need occupy only about as the technology evolves toward the production of larger one-fourth a square inch. and larger circuit functions on a single semiconductor On the silicon wafer currently used, usually an inch or substrate. For simple circuits, the cost per component is more in diameter, there is ample room for such a structure if nearly inversely proportional to the number of components, the components can be closely packed with no space wasted the result of the equivalent piece of semiconductor in for interconnection patterns. This is realistic, since efforts to the equivalent package containing more components. But achieve a level of complexity above the presently available as components are added, decreased yields more than integrated circuits are already under way using multilayer compensate for the increased complexity, tending to raise metallization patterns separated by dielectric lms. Such a the cost per component. Thus there is a minimum cost density of components can be achieved by present optical at any given time in the evolution of the technology. At techniques and does not require the more exotic techniques, present, it is reached when 50 components are used per such as electron beam operations, which are being studied circuit. But the minimum is rising rapidly while the entire to make even smaller structures. cost curve is falling (see graph). If we look ahead ve years, a plot of costs suggests that the minimum cost per VI. INCREASING YIELD THE component might be expected in circuits with about 1000 components per circuit (providing such circuit functions There is no fundamental obstacle to achieving device can be produced in moderate quantities). In 1970, the yields of 100%. At present, packaging costs so far exceed manufacturing cost per component can be expected to be the cost of the semiconductor structure itself that there is no only a tenth of the present cost. incentive to improve yields, but they can be raised as high The complexity for minimum component costs has in- as is economically justied. No barrier exists comparable creased at a rate of roughly a factor of two per year to the thermodynamic equilibrium considerations that often (see graph). Certainly over the short term this rate can be limit yields in chemical reactions; it is not even necessary expected to continue, if not to increase. Over the longer to do any fundamental research or to replace present term, the rate of increase is a bit more uncertain, although processes. Only the engineering effort is needed. there is no reason to believe it will not remain nearly In the early days of integrated circuitry, when yields were constant for at least ten years. That means by 1975, the extremely low, there was such incentive. Today ordinary number of components per integrated circuit for minimum integrated circuits are made with yields comparable with cost will be 65 000. those obtained for individual semiconductor devices. The I believe that such a large circuit can be built on a single same pattern will make larger arrays economical, if other wafer. considerations make such arrays desirable. MOORE: CRAMMING COMPONENTS ONTO INTEGRATED CIRCUITS 83
OCR for page 172
172 THE FUTURE OF COMPUTING PERFORMANCE Fig. 2. diagram to technological realization without any special engineering. It may prove to be more economical to build large systems out of smaller functions, which are separately pack- aged and interconnected. The availability of large functions, combined with functional design and construction, should allow the manufacturer of large systems to design and construct a considerable variety of equipment both rapidly and economically. IX. LINEAR CIRCUITRY Integration will not change linear systems as radically as digital systems. Still, a considerable degree of integration will be achieved with linear circuits. The lack of large- value capacitors and inductors is the greatest fundamental Fig. 3. limitation to integrated electronics in the linear area. By their very nature, such elements require the storage of energy in a volume. For high it is necessary that the VII. HEAT PROBLEM volume be large. The incompatibility of large volume and Will it be possible to remove the heat generated by tens integrated electronics is obvious from the terms themselves. of thousands of components in a single silicon chip? Certain resonance phenomena, such as those in piezoelec- If we could shrink the volume of a standard high- tric crystals, can be expected to have some applications for speed digital computer to that required for the components tuning functions, but inductors and capacitors will be with themselves, we would expect it to glow brightly with us for some time. present power dissipation. But it won’t happen with in- The integrated RF amplier of the future might well con- tegrated circuits. Since integrated electronic structures are sist of integrated stages of gain, giving high performance two dimensional, they have a surface available for cooling at minimum cost, interspersed with relatively large tuning close to each center of heat generation. In addition, power is elements. needed primarily to drive the various lines and capacitances Other linear functions will be changed considerably. The associated with the system. As long as a function is conned matching and tracking of similar components in integrated to a small area on a wafer, the amount of capacitance structures will allow the design of differential ampliers of which must be driven is distinctly limited. In fact, shrinking greatly improved performance. The use of thermal feedback dimensions on an integrated structure makes it possible to effects to stabilize integrated structures to a small fraction operate the structure at higher speed for the same power of a degree will allow the construction of oscillators with per unit area. crystal stability. Even in the microwave area, structures included in the VIII. DAY OF RECKONING denition of integrated electronics will become increasingly important. The ability to make and assemble components Clearly, we will be able to build such component- small compared with the wavelengths involved will allow crammed equipment. Next, we ask under what circum- the use of lumped parameter design, at least at the lower stances we should do it. The total cost of making a frequencies. It is difcult to predict at the present time particular system function must be minimized. To do so, just how extensive the invasion of the microwave area by we could amortize the engineering over several identical integrated electronics will be. The successful realization of items, or evolve exible techniques for the engineering of such items as phased-array antennas, for example, using a large functions so that no disproportionate expense need multiplicity of integrated microwave power sources, could be borne by a particular array. Perhaps newly devised completely revolutionize radar. design automation procedures could translate from logic 84 PROCEEDINGS OF THE IEEE, VOL. 86, NO. 1, JANUARY 1998
OCR for page 173
173 APPENDIX C G. E. Moore is one of the new breed of elec- tronic engineers, schooled in the physical sci- ences rather than in electronics. He earned a B.S. degree in chemistry from the University of Cal- ifornia and a Ph.D. degree in physical chemistry from the California Institute of Technology. He was one of the founders of Fairchild Semicon- ductor and has been Director of the research and development laboratories since 1959. MOORE: CRAMMING COMPONENTS ONTO INTEGRATED CIRCUITS 85