| document id: |
author: |
archive link |
publisher id: |
date written: |
| (notes for this file:) this document is part of "How to Build the Party of the Future" | ||||
Moore's Law will lead
|
Contents: TIP: Clicking on any of the paragraph numbers
--------- along the left margin
will take you back and forth
between the body of the article
and the table of contents.
paragraph
number chapters sections subheads
--------------------------------------------------------------
2 Appendix 1: Reference material:
internet use/ shrinking transistors
3 1. Estimates of internet use today:
17 2. Estimates of internet use tomorrow:
21 3. Estimated number of users on the internet
(1983-2001)
26 4. My own long-range estimates
of digital penetration:
34 5. Penetration of television (selected countries)
39 6. Microprocessors for PC's shipped last year:
44 7. The shrinking transistor:
50 How big is a transistor ?
58 8. Estimates of when Moore's Law will be dead:
62 Appendix 2: boddhisatva -- May 3
73 Appendix 3: Interview--Gordon Moore
107 boddhisatva -- May 26
122 Louis N Proyect -- May 26
|
Reference material: internet usage & shrinking transistors |
| (Numbers and percentages vary depending on who is counted and what qualifies as "internet use".) |
| Year by which more than 50 % of internet users will be located outside of the United States (ie: mainly Europe and Japan): (source: I can't remember) |
|
| Adult U.S. citizens who use at least one internet application besides e-mail: (source1) |
|
| U.S. adults who use e-mail only--or--who have tried to use the internet but have not stuck with it: (source1) |
|
| U.S. or Canada (16 years old and over) who have been on internet in any form in the past month: (source2) |
|
| E-mail and web users by frequency of usage: (source1) |
|
|
(source1: 1997 American Internet User Survey, NYT 5-7-97) (source2: March 1997 Neilsen Media Internet Demographics) |
| Note: I have not seen stats on this--but it is probably a safe guess that well over two-thirds of all internet usage is from businesses and schools. |
| Here are estimates by various analysts: |
|
| Don Tapscott is coauthor of "Paradigm Shift" and author of "The Digital Economy" (1996). Nicholas Negroponte is head of the Media Lab at MIT and also writes the back page editorial in each Month's "Wired". He is also the author of "Being Digital". Althou both Negroponte and Tapscott are very widely known and respected, this estimate of theirs has been widely ridiculed and my guess is that it is fairly unrealistic (see below). Michael Dertouzos heads the MIT Laboratory for Computer Science. His recent book on the digital communications revolution is titled "What Will Be" (1997). His estimates are more conservative and (in my opinion) likely to be more accurate. |
| (See my note below for a comment) |
|
| (Source: "The Digital Economy", Don Tapscott, Fig 1.2) (I estimated the number of users from Tapscott's logarithmic chart and hence this is a bit rough.) |
|
--- Note --- This chart is the likely source for the widely ridiculed estimate of one billion users by the end of the decade. The values for 1997 - 2001 seem to represent Tapscott's estimate based on the wildly erroneous assumption that exponential growth will continue unabated. By this logic, the number of internet users in 2010 will be one trillion and still going strong. Exponential growth may be characteristic of the early phase of the "S" curve representing the penetration rate of new technology, but this exponential growth must give way to more ordinary arithmetic growth as the various strata of the population which can afford the technology are saturated and new strata make purchases based on the declining cost and increasing usefulness. In spite of this kind of error, I highly recommend Tapscott's book "The Digital Economy". |
of digital penetration: |
|
Estimated penetration of digital infrastructure (prediction method: scientific wild-assed guess) |
|
|
| Above is my own estimate (from note 1.3 in POF-1) of the "S" curve likely to characterize the penetration of the many-to-many digital communications infrastructure. (I reformated it so that it should now be readable for readers who have mail programs set to proportional-spaced fonts.) |
| Note: there is no empirical data backing up this chart. Nor would any "think tank" or similar outfit dare offer such blatant speculation on how quickly the population of the world will be online. However I believe this graph is extremely important (if any reader believes there is a more important curve affecting the future of humanity--he is invited to offer suggestions of what it might be) and therefore made my own estimate. I have looked at the "S" curves for other communications technologies (phone, radio, TV, etc), done a lot of reading and, in the end, made a "scientific wild-assed guess". |
| Internet usage in many parts of the world will likely be a phenomena confined for a considerable time to the white-collar workforce. The driving force for a long time will be the need of corporations to make e-mail and the web available to their workforce in order to be competitive in the international market. Penetration of digital communications technologies to the consumer market--to the home--typically follows in the wake of wide-spread corporate adoption. In the gigantic U.S. consumer market this is already beginning to happen in a major way (PC sales in the U.S. are as big in the consumer market as they are to corporations) although the extraordinary size and wealth of the U.S. market makes it very atypical on a world scale. |
| Because a number of non-PC devices (such as the NC, or network computer) will emerge, and because PC prices are declining--it is still very unclear precisely what kinds of devices will emerge as the predominant type of many-to-many digital communications devices. In the U.S., set-top boxes such as "WebTV" (that convert TV's to e-mail clients and web browsers) are being sold for $250 with a $20/month fee for internet connection. Sales to date of such devices (under 60,000) are considered as being equivalent to zero in the massive U.S. market. Within ten years, however, all new TV's sold in the U.S. will be digital (ie: capable of being used as a computer display) and will likely have web browser ability, not as an option, but a standard feature. |
| percentage of total households that own a television |
|
| (source: Gallup India, WSJ 5-23-97) |
| The penetration of television gives a rough idea of what the penetration of digital communication devices would be once digital communication devices are as cheap as a TV. Note: statistics of telephone ownership would likely be a more useful and relevant indicator--because telephones require a monthly fee--but my apartment is so cluttered I cannot find these stats. The general order of widespread adoption of many-to-many digital communication devices, however, is indicated. China trails Europe, India trails China, and most of Africa (not shown) trails India. Penetration of telephones is considerably less than TV and is probably only 1 or 2 percent in India and Africa. More than half the world's population have never spoken over the telephone. Penetration of telephone (and possibly simple internet usage) in many of the poorer countries may initially be via wireless infrastructure since this is much cheaper, faster and easier to set up. |
| Microprocessors are the "brains" of a personal computer. The great majority of PC's have a single microprocessor. This number would be very close to the number of PC's sold. |
|
| (source: Economist -- 10 May 97) |
| PC's are currently the main platform for many-to-many digital communications. In the recent period, a large number of non-PC digital communication devices (ranging from boxes to set on top of a TV, to phones that send e-mail and browse the web, to fancy pagers) have either been brought to market or have been proposed. |
| Transistors are the switches that make digital communications possible. The number of transistors in a device are a rough measure of its computational "power" (ie: ability to do useful work) |
|
| Source: "As Limits on Chip Size and Price Near ..." M. Hiltzik (Los Angeles Times 5-12-97) (Sidebar) |
|
--- note 1 --- The great majority of the cost to produce a chip is the fixed cost to produce the first one (ie: R&D and building the plant to produce it). The variable cost for each additional chip is usually a small fraction of the price it sells for. |
|
--- note 2 --- The chart above showing the minimum feature size on a chip is very important but many readers may not have a clear picture of how big a micron is and its relationship to the size of a wavelength of light or an atom. Yet these relationships are important in order to understand the physical limitations which technology is now running into. Therefore I have added the following as background: |
| Most of us know how big a millimeter is. (For Americans unfamiliar with the metric system--there are about 25 millimeters in an inch.) There are a thousand microns in a millimeter. The diameter of a single human hair (about .003 inches) is approximately 75 microns. The wavelength of light is approximately half a micron. Therefore--while we may think of a wavelength of light as being very small--they are only about 150 times smaller than the width of a hair. |
| And, as we shall see, a transistor is about this size also, but just a little bit smaller. |
| There are a thousand nanometers in a micron. A wavelength of light would be about 700 nanometers for red and 400 nanometers for violet, with all the other colors in between. Hence, a transistor that is 0.35 microns in width--is 350 nanometers--and just a bit smaller than the smallest wavelength of light. This is why transistors can no longer be carved out by photons of visible light. Particles with shorter wavelengths are needed. |
| The portion of the spectrum with wavelengths shorter than visible light is called the ultraviolet. Ultraviolet wavelengths range from 400 nanometers down to 1 nanometer. Wavelengths smaller than 1 nanometer are called X rays. The diameter of a hydrogen atom (the smallest atom) is about a tenth of a nanometer. Atoms of silicon on a computer chip would be slightly larger and I will estimate here that there are about 6 silicon atoms per nanometer (the number 6 is only a guess but it makes the numbers come out nice and round). Hence a transistor that is 0.35 microns (or about 350 nanometers) represents a structure approximately 2000 atoms wide. |
| Hence, we could rewrite the chart above, using nanometers and atoms instead of microns: |
|
| When we picture transistor size in terms of atoms it is easier to gain an understanding of the kinds of difficulties that technology is encountering and why Moore's law (a law that states that the ability of technology to use light to carve silicon is improving resolution at a rate of approximately 20 percent a year) cannot go on forever. |
|
| Gordon Moore himself (see the interview below) says: |
|
boddhisatva's posts -- May 3 |
---------------------------------- Date: Sat, 3 May 97 4:27:03 EDT Subject: Re: M-I: (POF-1 Notes) |
| To whom..., |
| There is a company in southern California called Cymer Inc. which has made a tremendous amount of money because people were dumb enough to call an observation - that by Mr. Gordon Moore of Intel - a "law". Moore's "law", they reasoned, was so brilliantly predictive, that one need not diddle around with light lithography since the sainted Moore had predicted transistor packing only achievable through x-ray lithography. By this logic, Cymer's business - deep ultraviolet lasers for photolithography - was a loser. Well, of course Mr. Moore about as prescient as Kenny Kingston's Psychic Friends Network when it came to man's ability to overcome physics, and Cymer now has back-orders on deep UV light sources as far as the eye can see. |
| The moral to this story, one that should be painfully obvious to Marxists, is that "laws" which predict the end-product of human interaction well into the future are so much hokum. Yet, the temptation to make obvious and simple statements like "I think we're really gonna have to keep trying to pack those transistors on chips." into "laws" seems irresistible. The dynamic Mr. Moore hit upon: namely that processing speed begets the need for more processing speed, is certainly reasonable. Likewise, the dynamics Marx identified within society are equally reasonable and brilliantly observed, not because they are so complex, but because they are so basic. |
| We have to resist the urge to create arcane formulas out of fundamental observations about human relations. |
| peace |
----------------------------------------
Date: Sat, 3 May 97 4:39:31 EDT
Subject: Re: M-I: (POF-1) Appendix:
A Terror Discussed Nonstop
at Washington Cocktail Parties
|
| To whom..., |
| Of course, the fact that the Internet is a bourgeois phenomenon (at least for the foreseeable future) makes this even more interesting. Personal computers' having a penetration to 30% of American homes, with links to the internet less than that, with links to the non-AOL internet less than that (yet far above the average in the rest of the developed world) or accessible on a regular basis only to some majority proportion of college students, the arguments of Internet doom are arguments among elements of the bourgeoisie. That the Internet will have some bearing on the revolution is clear, but to what extent it will shape the views of the proletariat directly is not. |
| peace |
Interview with Gordon Moore |
| (The following appeared in the May 1997 issue of Wired. It is copyrighted and reproduced here for private use only.) |
| Moore's Law Repealed, Sort Of |
|
Gordon Moore foresees a day when his famous law breaks down -- well, maybe not. |
| By Peter Leyden |
| (sidebar:) |
| For more than 30 years, Moore's Law has governed Silicon Valley like an immutable force of nature. The idea that processing power will double every 18 months has been treated as an axiom -- rather than the rule of thumb it actually is. No one knows this better than Gordon E. Moore. In an obscure 1965 magazine article, Moore, then Fairchild Semiconductor R&D director, reluctantly predicted the expected increase in the power of integrated circuits over 10 years. By the late 1970s, Moore was a cofounder of Intel, and his tenuous "law" was well on its way to become a self-fulfilling prophecy among researchers, manufacturers, and vendors. Now, at 68, Moore will serve as Intel's board chair emeritus. Wired asked him to look to the next 30 years and, once again, make some predictions about the future of computing power. |
| (interview:) |
| Wired: How long will Moore's Law hold? |
| Moore: It'll go on for at least a few more generations of technology. Then, in about a decade, we're going to see a distinct slowing in the rate at which doubling occurs. I haven't tried to estimate what the rate will be, but it might be half as fast -- three years instead of eighteen months. |
| Wired: What will cause the slowdown? |
| Moore: We're running into a barrier that we've run up against several times before: the limits of optical lithography. We use light to print the patterns of circuits, and we're reaching a point where the wavelengths are getting into a range where you can't build lenses anymore. You have to switch to something like X rays. |
| Wired: Would X rays open a whole new round of doubling? |
| Moore: Theoretically, they keep us on this curve a longer time. Practically, they have a lot of problems. If we get away from optical lithography, somehow we have to get the subsequent technique up to the same level of sophistication to keep making progress rapidly. X rays represent a sufficiently dramatic change that it will be difficult to build on what we've done in the past. We'll have to start over, and it's going to take a long time to get traction. |
| Obviously, the industry is worried about this. We're looking at a US$200 billion industry that typically invests 10 percent of its revenues into research and development. A significant fraction of that will be aimed at solving this problem. Maybe something will come out that will make this transition a lot less onerous than I believe. |
| Wired: Are the costs getting prohibitive? |
| Moore: Recently someone gave me Moore's Second Law: The cost of manufacturing facilities doubles every generation. In the late 1980s, billion-dollar plants seemed like something a long way in the future. They seemed almost inconceivable. But now, Intel has two plants that will cost more than $2.5 billion apiece. |
| Wired: And the cost of each generation after that will double? |
| Moore: That's where you get into numbers that sound impossible again. If we double it for a couple of generations, we're looking at $10 billion plants. I don't think there's any industry in the world that builds $10 billion plants, although oil refineries probably come close. |
| Obviously our first reaction is to see what we can do to keep the technology moving but the costs down. For example, we used to build a completely new set of equipment each generation. Now our development people try to reutilize as much of the previous generation's equipment as possible. And they've been pretty successful. We may bring a $10 billion dollar plant down to the $5 billion range. But these are still huge numbers. |
| Wired: What will we be able to do with these superchips? |
| Moore: Even with the level of technology that we can extrapolate fairly easily -- a few more generations -- we can imagine putting a billion transistors on a chip. A billlion transistors is mind-boggling. Exploiting that level of technology, even if we get hung up at a mere billion transistors, could keep us busy for a century. |
| Wired: How much more powerful than today's chips are billion-transistor chips? |
| Moore: Our most advanced chips in design today, will have less than 10 million transistors. So, we're talking about a hundred times the complexity of today's chips. We wouldn't have the foggiest idea of what to do with a billion transistors right now, except to put more memory in a chip and speed it up. But as far as adding functionality, we don't know what can be done. |
| Wired: Do you think DNA computing, or organic semiconductors, could supersede microprocessors? |
| Moore: I'm skeptical about that stuff. You stir up a bunch of goo, and it's going to do something? I'm a chemist, so I can say this. The things we build don't happen that way. We're more deliberate in the way we do things. |
| I believe the technology our industry has developed -- this idea of building very complex structures layer by layer -- is a fundamental technology. It is as fundamental to the Digital Revolution as metalworking was to the Industrial Revolution. I don't believe it's going to be replaced. But I could be wrong; I could be too tied up in my own technology. |
| Wired: What about quantum computing or building computers with nanotechnology? |
| Moore: I'm skeptical about this, too, but it's closer to what we do than the DNA stuff. Quantum devices may be the ultimate transistors. The transistor doesn't behave very well when you get down to very small dimensions, but that gets into the realm where things like quantum devices start working. |
| We may make the transition to a kind of quantum device that keeps the whole trend going. Quantum devices are pretty far out, and a lot of work has to be done. They're far enough away that they're beyond my tenure in this industry -- a couple of decades from now. |
| Wired: Is there ever a point where you see so many problems on the horizon that you just want to give up? |
| Moore: Engineers thrive on problems. They're trained to solve problems. When they run out of problems, they become very frustrated. |
| Wired: I see. So you're just loving this. |
| Moore: Yeah, this is great stuff |
| ["It's as simple as that."] |