Slashdot Mirror
Andy Grove Says End Of Moore's Law At Hand
Jack William Bell writes "Intel chief Andy Grove says Moore's Law has reached its limit. Pointing to current leaks in modern chips, he says -- "Current is becoming a major factor and a limiter on how complex we can build chips," said Grove. He said the company' engineers "just can't get rid of" power leakage. -- But, of course, this only applies to semiconductor chips, there is no guarantee that some other technology will not take over and continue the march of smaller, cheaper and faster processors. I remember people saying stuff like this years ago before MOSFET." Update: 12/11 22:01 GMT by T : Correction: the text above originally mangled Andy Grove's name as "Andy Moore."
Shouldn't that be Andy Grove and Gordon Moore?
But, of course, this only applies to semiconductor chips, there is no guarantee that some other technology will not take over and continue the march of smaller, cheaper and faster processors.
I think they will just move away from silicon. Perhaps we have reached the limits of silicon but their is lots of research being done by acedamia and chip manufacturers on other materials.
FoundNews.com - get paid to blog.,
I'm curious what kind of results the experimentation in superconductivity and semi-conductors will yield. They sound kind of mutually exlusive. But we may yet see Moore's Law revived and revised...
Course, that's probably 15 years away...
...hearing this news the first time in 1989 and I read it the second time in 1994.
So. we'll see. I wonder if it now starts applying to graphic cards.
The industry is used to power leakage rates of up to fifteen per cent, but chips constructed of increasing numbers of transistors can suffer power leakage of up to 40 per cent said Grove.
:-)
No wonder my laptop only gets about a hour of runtime on its battery.
So, this means that anything that possibly can go wrong no longer will! Hey, I'm all for that!
What? Moore's Law? Oh. Nevermind.
Seeing as he is a big part of a major CPU firm Intel, is he being short-sighted (which I doubt) or is he trying to brace the market for a slowdown in CPU clock speed?
It might help the company if expectations for new CPUs aren't higher than what they can produce.
Personally, my vote goes for optical CPUs as the wave of the future. Larger than curent CPUs might not be a problem if they don't put off much heat.
Moderation: Put your hand inside the puppet head!
As the submitter eluded to; this has been said so many times before that I simply don't believe it. I remember reading the same thing about 100Mhz being the fastest we could build. Technology will find away as long as people are willing to buy it. And people will be willing to buy it because we all need to run Quake 4, 5, 6 etc.
The end of Moore's law is heralded on Slashdot every 2 months or so; it comes at the hand of new materials (copper, etc), new layering techniques, the ever-popular quantum computing, etc. Frankly, it doesn't seem to me to be that useful a benchmark anymore. The article says it will come sooner, but I foresee in 7 to 10 years the physical production, leakage stoppage and general quality of the chips will be so perfected that Moore's law will no longer be applicable to silicon chips. But, by then, new sorts of chips will be available to pick up the slack. So let us say farewell to silicon, and enjoy it while it lasts. It is like the fossil fuels problem really, except the industry is slightly more willing to advance, having set up years in advance a healthy pace to keep.
Now I can just buy a really fast computer and know that I'll never need to upgrade again!
I hope this means back to actually finding ways of optimizing code, and not the standard "We can throw money at it", or "Next year computers will be twice as fast".
However, may be better processor architectures and clusters will keep the march going.
Either way, I believe some progress would be made.
S
As long as Newton's law stays in effect I am not to worried.
:-) Sjaak
BTW do most of the users really need fast machines? I can do all my work without any problems on my 333Mhz PII
CU
I was waiting for the commemorative Pentium XT running at 4.77GHz.
Life is the leading cause of death in America.
Intel stock goes down like 50%
If it's the end, it wasn't a law to start with, then, was it?
Slashdot: Failed Car Analogies. Amateur Lawyering. Anecdote Battles.
I've always had issues with calling Moore's Law a "Law". Nobody has conclusively proven it. It should instead be called "Moore's Hypothesis" or "Moore's Theorem" if you're more optimistic...
Moore's law is finally coming to an end. Seriously, continous and rapid advance of processing power is the one thing that's holding back affordable universal and pervasive computing in schools. These cash strapped schools cannot afford to replace text books every two years, let alone computers that cost hundreds more. Things are better now because relatively useful computers can be had for very cheaply, compared to just a few years ago, but scrapping Moore's law altogether is even better. Steve Wazniak also agrees
In Soviet Russia, articles before post read *you*!
Oooh, so Mother Nature needs a favor?! Well maybe she should have thought of that when she was besetting us with droughts and floods and poison monkeys! Nature started the fight for survival, and now she wants to quit because she's losing. Well I say, "Hard cheese."
I write in my journal
..if Intel and AMD hadn't got locked into that stupid GHz battle and instead concentrated on optimizing their CPU design (rather than just ramping up the speed silly amounts) then there might have still be a few more years left before it became such a problem.
Maybe thats the way forward? Optimisations and improvements on the chips instead of raw clock speed....?
"Hey! Unless this is a nude love-in, get the hell off my property!!"
The number of stories posted on Slashdot about the end of Moore's Law will double every 18 months.
I think Silverman's Paradox applies here...
"If Murphy's Law can go wrong, it will."
Fascism starts when the efficiency of the government becomes more important than the rights of the people.
or am I wrong?
So we're running out of ways to pack more and more transistors into a device. There's still a ton of room to improve the layout of those transistors, the world is full of whines about x86 architecture.
This doesnt mean 'computers are as good as they're going to get', it just means the fabrication plants are as good as they're going to get.
I don't need no instructions to know how to rock!!!!
How many times do we have to hear people put their foot in their mouth? I would have thought Intel would've known better!
... is it good for?
But what
- Engineer at the Advanced Computing Systems Division of IBM, 1968, commenting on the microchip.
I think there is a world market for maybe five computers.
- Thomas Watson, chairman of IBM, 1943.
What can be more palpably absurd than the prospect held out of locomotives traveling twice as fast as stagecoaches?
- The Quarterly Review, England (March 1825)
The abolishment of pain in surgery is a chimera. It is absurd to go on seeking it. . . . Knife and pain are two words in surgery that must forever be associated in the consciousness of the patient.
- Dr. Alfred Velpeau (1839) French surgeon
Men might as well project a voyage to the Moon as attempt to employ steam navigation against the stormy North Atlantic Ocean.
- Dr. Dionysus Lardner (1838) Professor of Natural Philosophy and Astronomy, University College, London
The foolish idea of shooting at the moon is an example of the absurd length to which vicious specialization will carry scientists working in thought-tight compartments.
- A.W. Bickerton (1926) Professor of Physics and Chemistry, Canterbury College, New Zealand
[W]hen the Paris Exhibition closes electric light will close with it and no more be heard of.
- Erasmus Wilson (1878) Professor at Oxford University
Well informed people know it is impossible to transmit the voice over wires and that were it possible to do so, the thing would be of no practical value.
- Editorial in the Boston Post (1865)
That the automobile has practically reached the limit of its development is suggested by the fact that during the past year no improvements of a radical nature have been introduced.
- Scientific American, Jan. 2, 1909
Heavier-than-air flying machines are impossible.
- Lord Kelvin, ca. 1895, British mathematician and physicist
Radio has no future
- Lord Kelvin, ca. 1897.
While theoretically and technically television may be feasible, commercially and financially I consider it an impossibility, a development of which we need waste little time dreaming.
- Lee DeForest, 1926 (American radio pioneer)
There is not the slightest indication that [nuclear energy] will ever be obtainable. It would mean that the atom would have to be shattered at will.
- Albert Einstein, 1932.
Where a calculator on the ENIAC is equipped with 19,000 vacuum tubes and weighs 30 tons, computers in the future may have only 1,000 vacuum tubes and perhaps only weigh 1.5 tons.
- Popular Mechanics, March 1949.
(Try the laptop version!)
There is no need for any individual to have a computer in their home.
- Ken Olson, 1977, President, Digital Equipment Corp.
I have traveled the length and breadth of this country and talked with the best people, and I can assure you that data processing is a fad that won't lastout the year.
- The editor in charge of business books for Prentice Hall, 1957.
[Quotes from this page.]
If you celebrate Xmas, befriend me (538
Ya, I mistyped. Slips happen.
- -
Are you an SF Fan? Are you a Tru-Fan?
A law is a law...and it's time corporations were held responsible.
I expect the Feds to start handing out stiff penalties to processor manufacturers who fail to meet the law's demands.
Karma: Excer..ex...excellahhh...realll good (mostly affected by drinking not done in moderation)
However, if you define Moore's law as computational capacity doubling every 18 months, than it is very unlikely to end. If you project back to well before integrated circuits, or the law itself, computational capacity has been growing at this same exponential rate for many decades - even back to the earliest mechanical based "computers". There will be something to replace the current paradigm; the paradigm has already changed numerous times without throwing off the exponential curve.
For a facinating look at this phenomenon at what it holds for the future, I'd recommend The Age of Spiritual Machines: When Computers Exceed Human Intelligence by Ray Kurzweil.
Random is the New Order.
Is an economic law, not a physical one. Lack of demand for high-powered processors is going to slow the progression in processor speeds.
love is just extroverted narcissism
That's about right. It's a bit more pessimistic than the SIA roadmap, but it's close. Grove was just stating, for a general audience, what's accepted in the semiconductor industry. Optical lithography on flat silicon comes to the end of its run within a decade. Around that point, atoms are too big, and there aren't enough electrons in each gate.
There's been a question of whether the limits of fabrication or the limits of device physics would be reached first. Grove apparently thinks the device problem dominates, since he's talking about leakage current. As density goes up, voltage has to go down, and current goes up. A Pentium 4 draws upwards of 30 amps at 1.2 volts. We're headed for hundreds of amps. It's hard to escape resistive losses with currents like that.
There are various other technologies that could lead to higher densities. But none of them are as cheap on a per-gate basis.
I attended a talk some 1.5 years ago by guy from Philips NatLab (home of the CD), which was called "Mo(o)re or less?". Although the talk was extremely boring and i forgot the final conclusions i do remember some potential showstoppers he listed:
-Of course the ultimate limit of a 1 atom transistor, can't remember the date this would occur
-Limited speed of signals acros the chip: If the clock frequency gets much larger a signal would require several buffer stages to reach the other side.
-Capacity of wires gets more important: the interconnects don't scale at the same pace as the transistors. Their finite capacity limits clock speeds
Some non-technical reasons:
-Increasing costs of new fabrication processes: each new increment is more expensive.
-Limited manpower to design circuits with more and more transistors. This probably means that a larger area of the chips will consist of 'dumb' circuits like cache.
karma police: arrest this man, he talks in maths; he buzzes like a fridge, he's like a detuned radio. [radiohead]
Moore's law hasn't reached any limits, we have. If this is a barrier we need to overcome, we will overcome it. We could be be thousands of years ahead of our time in our technology if that was our priority as a race, or even as individual nations. If we *needed* faster, smaller processors, the governement would pour money into R&D and more brilliant minds could be gathered to work cooperatively and the results would be results :)
Seriously, we've risen above much greater challenges than this..
It sorta sounds like Intel is about ready to quit trying to innovate, perhaps this is time for AMD to take the lead..
Everyone is entitled to their own opinion. It's just that yours is stupid.
Richard Feynman's address to the American Physical Society is a good intro to the physical limitations of miniaturization as it applies to Moore's Law. Also intersting, is the Law from the Horse's mouth found on this Intel page.
Logic is not Divine.
The power is largely dissipated as heat. [emphasis added]
Duh! Funny, I have never seen any (properly connected) microprocessor chip generating much in the way of light , sound, or X-rays. I suppose a teensy weensy amount might go off as RF emissions, but not from the DC leakage current.
Recall, AMD just said they are done trying to up clock speeds all the time. Now Intel is outting themselves, too. The fact that these companies are not saying things like "we need to go to other materials to get higher clock speeds" is because 1) it costs huge $$$ to research and develop new materials, 2) it costs serious $$$ to change fabs to use new materials, 3) NOBODY (no, not even you) wants to continue to pay for increased clocks when there is almost zero benefit in real applications.
Moore's Law is not dead. What is dead is the need for Moore's Law. I am not alone in noticing that, after 20 years of regular performance increases, things are now pretty good on the desktop, and excellent in the server room. Real changes now need to be in software and services. Further, high-performance computing is going the route of multiple cores per CPU, multiple CPUs per box, and clusters of boxes. The latter is probably the biggest innovation since Ethernet. So, who needs Moore's Law?
Intel and AMD know *all* this. They want out of the clock race, and yesterday. They want to get into the next level of things, which is defining services and uses for their existing products. They are seeing the end of the glamour years of the CPU and the rise of the era of information applicances, which *must* be portable. Users *will* be far more sensitive to battery life and perceptions of performance (latency and ease of use) and far less sensitive to theoretical performance measures.
Flame me if you like, but the geek appeal of personal computers is disappearing. Sure there will be people who fiddle with CPUs as a hobby, just as they did 30 years ago when the Apple computer was born to serve a small group of hobbyists. But is that the mainstream? Is that going to support Intel and AMD in their race? Are those companies going to promote a revolution in fab technology, to the tune of half a trillion dollars in investment and technology between them, just to support geeky hobbyists? They could, but they won't, because that is not the future. It is the past.
The future will still be interesting, mind you, but the challenge has changed. A phone that fits in your rear molar and runs off chemical reactions with your own saliva looks far more lucrative to these companies than a CPU that runs at 100Ghz and consumes as much power as an appartment complex.
=^..^= all your rodent are belong to us
You can fairly easily raise the threshold voltage (for a process). It makes the chip slower, but leaks less current (and therefore usually uses less power). This is one of the key elements of "Low Power" processes like CL013LP.
For more information, the Britney Spears' Guide to Semiconductor Physics is sure to help.
Interestingly, Using leaky transistors that switch faster has been a trick used for a very long time. One of the reasons the Cray computers took so much cooling was that they didn't use MOSFETs, their whole process was based on PNP and NPN junction transistors. For those who don't know much about transistors, FETs (or Field Effect Transistors) make a little capacitor that when you charge it up (or don't charge it up, depending), it lets current flow through on the other side. It takes a while to charge up the capacitor (time constant proportional to Resistance times Capacitance, remember!), but once it's charged there isn't any current (except the leakage current) that flows through.
At least, that's what I recall from my classes. I didn't do so well in the device physics and components classes.
-- Erich
Slashdot reader since 1997
I guess algorithm analysis will at some point become more mainstream again. I suppose application profiling will also become more popular.
Interestingly, the available memory will continue to grow, so we might end up structuring our data structures so that access time will be minimal. That is - our data structures will continue to change focus from compactness to raw speed. And big O analysis is part of that picture.
I think we'll see some interesting things happen with fiber technology, though. When those envisioned optimal silicone chips become commonplace and thus really cheap, all appliances might run on them, and thus make it feasible to distribute your processing between your computer, your fridge and your iron. We'll just interconnect everything - perhaps a new fibre connector in our electricity plugs.
Stop the brainwash
I thought the root of Moore's Law wasn't the technology involved but the drive for improvement in computation. So that the chips may not improve beyond a certain point but then making a massively parallel system on a 2"x 2" card would still go into Moore's Law. It is hardware independent.
I'd never put a limitation on this since somebody's going to come up with an idea to eek out more clocks.
What is music when you despise all sound?
We've now reached the stage where handheld devices have the same sort of processing power and memory of respectable desktops of a few years back, and I find it interesting that the sudden big hype is the tablet PC, which is relatively low speed but has good battery life. That could be the direction things are going, and if so it is hardly surprising Andy Grove is worried about leaking electrons, what with Transmeta, Via and Motorola/IBM having lower power designs.
A case in point about technology demonstrators. Someone mentioned aircraft. OK, how much faster have cars got since, say, 1904 when (I think) RR first appeared? Not an awful lot, actually. They are vastly more reliable, waterproof, use less fuel, handle better, are safer, and enormously cheaper in real terms BUT they go about the same speed from A to B and carry about as many people. And they are still made of steel and aluminum, basically the same stuff available in 1904.
This is far from a perfect analogy because, of course, the function of the computer keeps getting reinvented: it is applied to more and more jobs as it gets cheaper, more powerful, and more reliable. But it does point out that the end of Moore's law is not the end of research and development.
Panurge has posted for the last time. Thanks for the positive moderations.
Just enclose the processor in a static warp field and adjust the speed of light in your new proto-universe. Sheesh, come on people.
What if it is just turtles all the way down?
Colloquially we speak of Moore's Law and we mean "Chips get twice as fast every 18 months."
This is not what Gordon Moore said. Moore's statement was based on transistor density. Indeed, perhaps we may not be able to cram transistors together as much in the not too distant future.
Does this mean that chips won't continue to get twice as fast every 18 months? It would surprise me if processors slowed down their rate of speed growth much this decade. As people begin playing with digital video on the desktop, as people write games that can actually push enough information to a GeForce4 FX to make it worth spending money on, people are still going to want faster and faster machines. And while AMD still exists as a competitor to Intel, even those people who don't really need a 7 GHz machine are going to find that that's what's available.
So while Moore's law, as it was stated, may be nearing its end, Moore's law, as it is usually spoken will probably stick around for a good while longer.
Honestly, I think a bigger trend will be to take advantage of formalisms that let developers develop more reliable and stable software. Now, I know and you know that things like functional programming have been out there for years, and haven't succeeded because first, they were too slow and therefore wasted too many processor cycles. This is obviously much less of a problem now - Java "wastes" lots of processor cycles, but for a lot of software needs, saves so many human "thinking" cycles that it pays off in spades for businesses that need business or enterprise software to Do Stuff for the back-end sides of industry.
So what big problem(s) are left in the software world? Well, people still bitch about how fucking unreliable most software is. In particular, core, critical system areas, like the interface between hardware and software - as more hardware is out there, and more drivers are developed, and backwards compatibility is an issue, hardware interactions have not become substantially more reliable. And frankly a lot of applications themselves, have become substantially less reliable - the big problem is that adding features and changing GUIs seems to break too many things and introduce too many potential problems (look at Outlook XP vs. Outlook 2000 - fixed some security holes, made a prettier GUI, and made the damn thing crash all the time).
Look at a lot of the academic work being done in computer science, especially in programming language design, operating system design, parallel algorithms and parallel languages. Sometimes researchers head off down dead-end paths, but sometimes they have it right, and it just takes a while for industry to see what they need this stuff for. That being said, it'll always be cheaper to teach people "Programming in Java 101" in India and then hire 1000 of them to hack away at code, admitted usually for the most uninteresting and repetitive types of development work (at least, this will hold until economic parity in the third world becomes a reality).
...what this means in relation to Intel's .09 micron work with Prescott (slated for late next year)? Could be nothing, could be indicative of INtel hitting some stone walls in .09 micron development (which I always knew would be a tough row to hoe for complex cpus.)
Read one post earlier in which the poster thought AMD was abdicating a "clock speed" race. Obviously, this sentiment, among so many like it, comes from Hector Ruiz's speech last week in which he said that AMD wasn't going to do "technology for technology's sake." I wish Hector had made himself a bit clearer...;)
What I think he meant was that unlike Intel with Itanium, AMD was not going to design brand-new technologies with no practical worth simply for the sake of performance (because Itanium has no software it's very nearly useless--except for doing PR benchmarks for Intel.) That's why AMD chose to do x86-64--because it is technology for practicality's sake. That's my take on that statement.
Also, AMD has been out of the "clock race" ever since they designed the K7. The race AMD wants to win, and has been winning, is the "performance race" which doesn't depend on raw MHz. Any P4 will be much slower than any K7, when clocked at the same MHz speed. That's why AMD's been using performance ratings--because they are much better measures of performance than mere MHz speeds could ever be between competing cpus with differing architectures.
Regarding the natural world environment, you're correct, as I've seen some harsh criticism of the volume and toxicity of waste byproduct of semiconductor manufacturing. It's not so simple as, just add a little sand and some magic and voila! It's probably not reported so much because the wonders of innovation and heated competition make for more sexy news writing.
Something not mentioned much, but observed by more than a few grumbling parties, is the ever increasing size of code. My first encounter with this was upgrading from RSTS/E 7.? to 8.0, which was darn exciting back in the day, yet the size of the kernel would have been about 50% larger if we activated all the features *I* wanted to (and since I was the admin, lemme tellya, it was darn painful to trim off a few features I lusted after to squeeze it into our memory and performance target.) These days, it's often the OS, ever notice how Windows installs got to needing more space than your entire first harddisk? Common response seems to be, just throw more memory at it. Yet, I think there's a Moore-like law with versions of Windows, i.e. every 2 years a new version comes out with twice as much code.
With physical limitation of the current components nearing the top of the "rate of declinging return" curve, poor performance of the software will eventually catch up with users expectations. Thus, leaner, faster code could become a market direction.
"** NEW: Office-O-Lux, With 50% less redundant code! ***"
A feeling of having made the same mistake before: Deja Foobar
Eventually you will reach a limit on the size of the individual swtiches. The one the article gripes about appears to be the sloppy wave function of the electrons letting them tunnel across the junction. But matter is lumpy (quantized) and eventually you'll hit a just-a-few-atoms wall.
... Keep this up until you are done. Laying out your gates for minimum signal run length means you end up with a cube, or something close to it.
But there's more that can be done - in terms of geometry and organization.
Current chips are a single two-dimensional array of components (or sometimes a small number of layers). But build your gates and interconnects in 3-D and you can go farther on two fronts:
- Speeding up the individual functions a bit further. (The more complex, the more improvement).
- Combining a LARGE nubmer of parallel elements into a small space (so they can talk to each other quickly).
Back in the '70s I had a rap I'd do called "preposterous scale integration". Basic idea:
- Use diamond for the semiconducting material (because it conducts heat VERY well).
- Grow a LARGE sheet of it, writing the domain doping and interconnects with ion beams as you go.
- TEST the components as you go:
- Negative power lead is a slow (low accelleration voltage) electron beam.
- Positive power lead is a fast (high accelleration voltage beam) electron beam, causing secondary emission of more electrons than are in the beam.
- Test injection probes are smaller versions of the power leads.
- Test probe is a very slow electron beam, where the electrons turn around at the surface, and a positively-charged region will suck 'em to the chip.
(These are all variants of electron microscope imaging hacks that were in use as far back as the 70s.)
- If a component fails, turn up the current, vaporize it, and deposit it again. Repeat until you have a good one.
- When you're done with the layer, don't stop. Deposit another layer, and another,
- Apply power to two opposite faces of the cube. Use bus bars the size of the cube face - at least near the contact point - to minimize IR drop. Use a good conductor, like copper or silver.
- You need a LOT of cooling. So circulate cooling liquid in the buss bars. (Copper and silver are also good heat conductors, and water is a terrific heat carrier.)
- The other four faces are for I/O. Use semiconductor lasers, photodiodes, and fiber optics light-pipes. You can COVER the faces with fibers. Put your drive electronics and SerDeses in the layer just under the pipes - or dope the index of refraction of the diamond to make a light-pipe into the depths and distribute them throughout the volume.
- Diamond is stable up to very high temperatures, but you need to protect it from air when it gets hot (or it will burn). So put it in a bottle with an inert gas just in case. Limitiing temperature structurally is about where it starts going over into graphite, so you can let it get up to a dull red glow (if your I/O is at some bluer color and that temperature doesn't create too much thermal noise).
- How big can you get? Square-cube law limits your I/O-to-computation ratio, since the I/O is on four faces that go with the square of the linear dimension, the computation goes (approximately) with the volume, or the cube of the dimension. The cooling-to-gate ratio suffers a similar square-cube issue (plus a linear penalty for power losses from the internal distribution busses). You also have an interconnect penalty - as you get bigger you have to give a higher fraction of your volume to power and signal lines (or signal repeaters), but this actually improves the square-cube problems. Finally, construction time is about proportional to number of computational elements. So let's pull a number out of nowhere and say two meters on a side.
Of course the punch line is what the device would look like:
- A six-foot cube of diamond.
- Glowing cherry red.
- In a glass bottle of inert gas.
- Supported by water-cooled silver bus bars.
- And connected to everything else by an enormous number of glass fiber light-pipes.
In other words, the kind of thing you'd expect to be the ship's brain in a late model Sklyark spacecraft, from one of George O. Smith's golden-age science fiction novels. B-)
====
This rap was always entertainment rather than a serious proposal, and is no doubt buggy. For instance: I hear doping diamond is a bit problematic. And these days I'd suggest doing chip-under-construction powering and testing using physical contacts and JTAG fullscan or a variant of the CrossCheck array, rather than (or to suplement) the electron beams.
But I hope the point is made that, for parallizable tasks at least, we still have a LONG way to go with improved geometry before we finally hit the wall.
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
It says that either processor speed (or density) will double every 18 months OR (and it's a big or) the price will halve in 18 months.
So logically we could continue on with the same speed processor and just have them get progressively cheaper. But hmm, I wonder whose profit margins this would affect? What he's setting us up for is that Intel will refuse to lower their prices. They'll continue to make the chips cheaper and cheaper but they won't sell them for any less.
I actually look forward to an end in ever increasing clock rates, because then we can all get back to programming skillfully and making tight efficient code.
--------- Beware the dragon, for you are crunchy and good with ketchup.
IBM has been using partially-depleted SOI which actually increases leakage current and therefore increases standby power.
Fully-depleted SOI should have lower leakage current due to better control over the transistor channel. While Intel doesn't call it SOI, they announced their "terahertz transistor" sometime last year which is actually a fully-depleted SOI device.
Another way to reduce leakage power would be to use dual-gates when building the transistor. There is a decent amount of research going on in this field. Dual gate would offer large decreases in leakage current.
You have to remember the chips improve with time. More importantly things other things that effect speed also improve quite a bit. For example when the 386-20s came out there weren't caches so the chips ended up pulling no ops extremely frequently.
Anyway taking your comparison and using a benchmark of the time (the Norton System info benchmark):
80286-16 got a 9.9 (i.e. 9.9x as fast as the XT)
80386-20 got a 17.5
More importantly the cache configurations that came with the 80386-25 raised the score to a 26.7
adjusting for the increase in mhz:
26.7 * 16 / 25 = 17 which is close to double.
I'll stand by my statement.