Slashdot Mirror
Gordon Moore On Moore's Law
missingmatterboy writes: "Technology Review has a wide-ranging interview with Gordon Moore, wherein he discusses the future of computers, his famous 'Moore's Law,' the need for better education, the environment, and finally, why he, along with Bill Hewlett and Dave Packard, picked up the tab for SETI. Cool guy." Who better to ask about the future?
Selective memory, perhaps? AMD announced the 1GHz Athlon mere hours before Intel announced the Penium III 1GHz. It's a shame that such a lame bunch of partisan "misrememberance" is rated so highly.
Linux user since early January 1992.
My cock is bigger than your cock.
With all due respect for one of the true giants of the semiconductor and IT industries, but I don't think he has to be the best man to ask about the future. Not because he says so himself (for those who didn't read the article: he does), but because he's been involved in it for so long already.
In general, the fate of all experts (especially in a world moving as fast as ours does nowadays) is that, with aging, they inevitably get to a point where they don't see the changes coming anymore. They end up relying more and more on extrapolations of their own (sometimes vast, but still very limited) knowledge and experience. At some point these extrapolations break down completely. IMHO, it's a rare person who understands this to be applicable to himself or herself in time. Those who do, stop making predictions years into the future anyway.
--
Linux user since early January 1992.
This article says that the Moore's Law says the number of transistors will double every two years. I always heard it was every 18 months.
That was metric years.Trolling is a art,
I once read an article (or was it a chapter of a book? Can't remember now) discussing the infamous "Drake Equation" (regarding the probability of detecting intelligent life via SETI).
The author had pointed out that the variable intended to represent how long a form of intelligent life existed before dying off really meant "how long a form of intelligent life broadcasts detectably in the radio spectrum", as far as SETI is concerned.
The author then speculated that strong, readily detectable radio signals from Earth will have been going roughly 100 years before cable, fiberoptics, and other "non-broadcast" means of communication start supplanting them.
He then plugged THAT number into the Drake Equation and got...1.
"That must be us." the author quipped...
Not that I have a problem with SETI or anything, but I found the argument very interesting...
---
Hacker Public Radio is our Friend
The .13 micron processes that are state of the art right now are actually pretty big compared to what would be required in a few years, at least by Moore's Law. The problem is that technologies are lagging behind the "Law". A prime example is in lithography. Commercially practical sub-.1 micron lithography doesn't exist. Extreme UV hopes to drive device sizes down to as low as .04 microns, but it's still very experimental, even in its 4th year of development.
Not to belittle Gordon Moore and his "Law", but I think that it's about to give out. Of course, what we call Moore's Law was really nothing more than an off the cuff remark by an "important person", so by following it, the semiconductor industry has validated it.
Anyway, I don't think that it matters whether or not the industry follows it...after all, what we're really after is faster, better devices. And if it's possible to get there without following Moore's Law, then what's the difference? I think that's where we're headed.
-h-
Thank god I own AMD...
--
Poliglut
Yes, how odd. Someone who probably knows an order of mangnitude more about the Kyoto treaty than you claiming that Bush is on the pragmatic end of protecting the environment and probbably will be OK, while you on the other had are pretty sure Bush would be willing to nuke all national parks from orbit, if only he had the nukes to spare.
I wonder who has the viewpoint that people should pay "moore" attention to? Happily for you around here it's sure not Moore!
"There is more worth loving than we have strength to love." - Brian Jay Stanley
I think that we're pretty low on the curve of being able to find stuff in general, let alone inteligent life. I for one loose my keys and my wallet almost every day for at period of about five minutes and then there's my cell phone for which I sometimes conduct near-exhaustive searches of my domain only to find that it is in my pocket. I'm not the brightest but I know that there are a shitload of people dumber than me (e.g. Clinton/Gore supporters and the 70-odd percent of the population that believes in creationism instead of darwinsim). Granted some of the brightest minds on the planet are doing the searching but the fact that they rely on us--the ignorant masses--for their survival shows pretty clearly that they're no even near optimal vs. our actual capacity to conduct the search.
If the life is truely intelligent then it is avioding us because lord knows we're not too friendly. If you are an inteligent species outside of the planet earth and less powerful than us we'll probably conquer you and if you're more powerful then we'll do whatever we can to become more powerful so that we may later conquer you. Why do I think so? Well, that's pretty much been our modus operandus for all of recorded history.
Then there's the issue of what we can reliably detect. We do know for certian that every one of our ideas about physics basically only works in the domain we can easily observe and even sometimes in our observable domain we don't have good ways to explain what the hell is going on. Add the problem that we're pretty sure that most of the universe consists of dark matter and dark energy that we can't even detect yet and you get the idea that as our gaze wonders past the limits of our biosphere our quality of perception decreases (i'm guessing!) exponentially.
There's some major drawbacks to our existance that I think will stunt our ability to overcome our myopia. One is the fact that we're trapped in linear time and we can only traverse our time in a single direction at a fixed rate. Also we live in only 3 physical dimentions. While this 3D existance seems pretty advanced to us (ask SGI or Nvidia) there's plenty of scientists out there much smarter than I am who are embarassed to try to even explain how we're missing out on the action in the higher echelons of dimentionality (why?--even the small portion of us who even accept the idea that there are more dimentions tend to forget about it pretty easily which is illustrated by the death-rate for passengers of motor-vehicles.)
Do I think we should give up because we won't find anything anyhow? Hell no! We aught to pour all our money and resources into searching, colonizing and exploring outer space. We've got nuclear weapons, acid rain, rising sea levels and major league baseball. That is proof enough for me that this planet is fucked and it really is time to move on. I, for one, have a lot of new and exciting ideas for how we can screw up other habitats and no one is going to let me try that stuff around here. I want to move on.
If you haven't noticed the USA is the most powerful and free country in the world because it was created by people who habitually run away from civilization because they would rather try to survive alone in an (relatively) unknown wilderness than live with the schmuks they grew up with. Did I mention that a lot of the time the wilderness to which they ran posed significant threats of death, dismemberment or even boredom. Well guess what? There's no more wilderness! We filled it with schmuks! The implication here is that we're going to have either learn to live with eachother or devolve into a society similar to the europeans and probably become, like them, a bunch of socialist wimps with traditions and culture and all that other garbage. We need space! It's our final frontier!
I will give everything I own and hold dear to anyone who can get me off this rock to someplace where I can be away from the rest of these jerks--even if there is significant danger of boredom! Long live NASA. Long live SETI (but not too long because it goddamned better start working soon or i'm going to have to detonate a nuclear device). Cordially, Rob
-jhp
/. -- the Free Republic of technology.
Along these lines, you might want to check out a scientist bold enough to consider the non-3+1-dimension alternatives and attempt to explain why they wouldn't 'work' for universes containing observers. I wouldn't consider it proof, but I found it to be a cool provocative paper explaining perhaps why we have 3 spacial and 1 temporal dimension. --LP
According to the Drake equation, the number of communicable civilisations in the galaxy increases by one per year of our looking.
Explanation here
--
Consultancy: If you're not part of the solution, there's money to be made in prolonging the problem
While this is true, I suspect the P3 1ghz announcement came to be in the following senario...
(Fade into Intel's Marketing Department)
Marketing Drone 1: Holy crap! AMD is set to announce their gigahertz Athlon processor. We MUST announce our own.
Marketing Drone 2: I'll call down to the engineers and see what we've got to match AMD.
(on phone)
Engineer: What?!?! We've just started making 600mhz chips. You're f-ing out of your mind!
Marketing Drone 2: But we must keep up with AMD! Can you build gigahertz processors?
Engineer: I'd estimate that about 0.5% of our current chip yield *could* make 1 gigahertz.
Marketing Drone 2: Great! I'm writing the press release right now. Get the bunny people up here!
Meanwhile back at the ranch...
Why do I think this is true? The P3 1ghz chip was very unstable compared to their other chips. They needed to write new BIOS code that disabled most of the preformance enhancing features just to get the chip to run. No, this wasn't misrepresentation, just the cold hard facts.
In this interview Moore mention that the chief reason that more people are not entering into the technical field is that kids are not really properly introduced to it in thier pre-college years. I fully agree with this belief, but i also think it is due to some other factors. Although a nerd's status in society has greatly improved since the eighties, there still exists a stereotype that nerds(hackers, engineers...) are socially inept and romantically unatractive. This belief is instilled in most children, forever deterrering them from a deep interest in computers. To some degree of course this stereotype is true, but then again most people who have highly developed social skills and popular personalities are the ones who tend to avoid the technical fields.
A rabbit in the hand is worth 4 in the cage
Intel became brainwashed into thinking that Moore's Law was an actual law, instead of a generally observed trend over the past several years. Yes, it has been quite a sustained trend, but still... not a Law in the sense of a proper Law.
Intel's problem was that they built their company around this "Law" and were surprised when little AMD didn't seem to care that they were going ahead of Moore's schedule! And Intel had so many plants around the world that it is taking them until now to upgrade them to the point where they should have been a long time ago to maintain technological leadership.
I remember reading last year that when AMD was beating the pants off of them with the 1G Athlon, Intel was still revved for producing P3-600's, more in line with Moore's law.
-rt-
-rt-
** Evil Canadians are taking over the world. Learn about the conspiracy
But when the system is constructed with thousands of individual PCs, which get upgraded anyway every few years, the entire computer gets a speed boost without having to be completely redesigned and rebuilt. I'd be curious to see how well the speed increase of SETI@home has matched Moore's Law over the two years since its conception, I bet it is pretty close.
Moore's comments on the education system in america strike home with me preety hard.
You almost cannot learn what you need as a base for a technical career in a average school enviroment. With all the cut backs and distractions in schools these days the future geek of america are gonna keep sucking as a majority of those working in the industry right now. How many of us know a moron working as an admin? Heh how many of us are morons?
once again we can see america getting flushed, and as far as I'm concerned at this point, we deserve it.
Guttermouth is a really good band.
Er... Well, y'know. You can't make an omelette without um... destroying a forest. Or something.
My only political goal is to see to it that no political party achieves its goals.
It's refreshing to see someone share the wealth like Gordon Moore does. These days, it seems that the accountants decide the fate of speculative research--How silly, especially when nearly all revolutionizing discoveries throughout history were due to sheer accidents or wacked-out ideas. Open letter to all you *illionaires: you can't take it with you!
I think /. should interview him. I would like to find out what he thinks about psychedelics.
Ewige Blumenkraft!
Ewige Blumenkraft!
I think it would be cool to give $5,000-20,000 grants to masters and doctorate students to finish their research, and make something useful out of it. If other people find it useful, then a company can buy it and market it. With $100 mil to give away, they can certainly afford the infrastructure.
Hmm .. I seem to remember this topic being discussed before (hasn't everything?) but I can't seem to find it..
-B
Not that this wasn't entirely predictable.
Weired, this Mooore.
He donates half of his Intel shares ($5 Billion) to environmental & educational programs, and on the other hand he sat on Bush's environmental advisory committee and does not even criticise the guy...(and we all know what Bush has done to the environment, including getting rid of the Kyoto agreements and licking every oil company's arse.)
I suppose he's just confused.
Still, I kind of like the guy giving away 50% of his fortune (wouldn't see old $Bill doing that!)
--
yours ever, fz.
I think people generally forget that "Moore's Law" was an off the cuff optimistic guess made in an interview with Electronics Magazine in 1965. He was pretty lucky with his guess, which has held generally true for the last 35 years. The guess of exponential growth fit in really nicely with the growing industry, and emergence of the microchip in everyday life. Greater demand meant greater research, and lower prices, spurring even greater demand. However, I see a set of different factors greatly slowing that cycle. First, as Moore said in the interview, the limits of the medium. Sub-atomic transistors are still pretty far on the horizon, and our current designs can't get much smaller. Second, a factor Moore didn't mention, the eventual market saturation. There is no radical new application on the horizon for the microchip to cause its continued spread at such a great rate. Sure, we all like to get faster, and faster computers, but most people already have somthing that works. Unless a new market opens, investment will slow, as will research, and the entire cycle will slow.
My Karma is so good, I'm the Dalai Lama...or something.
So more made a law.... then he changed it... now he's changed it again (but, made it so the law changes on it's own?) [intresting idea]
so.... This is how it goes
It every 2 years the number of tranistors in chips will double
Every 10 years that number (e.g. 2) will double.
Every 15 years Moore will change his law again.
--- My Karma is bigger than your...
------ This sentence no verb
Sorry. Didn't you read the fine print?
"Past performance is not a guarantee of future results."
main(c,r){for(r=32;r;) printf(++c>31?c=!r--,"\n":c<r?" ":~c&r?" `":" #");}
You have a very good point. Double the processing power= double the amount of ETI reports we can go through, but there is another law that will keep the chances low: lightspeed. 186,282 Miles Per Second. The closest star is 4.3 light years... 8.6 years for a round trip. Then you take the chances of a planet existing around Alpha Centauri (low), take the chances that the planet orbits in the life zone (which is relatively thin), then take the chances that life would evolve, then take the chances that the life would become intelligent, then take the chances that that intelligent life would live long enough to develop radio, then take the chances that the ETI would be seaching the particular radio band we are broadcasting on... and so on. Then apply that for planets thousands or millions of lightyears away, and you see that processing speed is but a minor factor in the search for ETI. As I said, you did bring up a good point, though.
Elmo knows where you live!
This article says that the Moore's Law says the number of transistors will double every two years. I always heard it was every 18 months.
Tyler
If the number of transistors on a chip doubles every two years, does that mean that our chances of finding extraterrestrial life also double every two years?
Tyler
The number of idiots using computers is directly proportionate to the number of AOL CDs
TODO: Something witty here...
If you look back to the early days of computer development (circa 1940s) to current technology and extrapolate the amount of computing cycles/$1000 ---- You will find that Moore's Law holds even before silicon. Therefore, it is reasonable to assume that Moore's Law will hold even after we have reached the limits of silicon technology (predicted within the next 15 years).
As below, so above and beyond, I imagine drawn beyond the lines of reason. Push the envelope. Watch it bend.
The great physicist Enrico Fermi raised the following objection to the existence of extraterrestrial intelligence (as distinct from extraterrestrial life):
It should take about 1-10 million years for an intelligent race to colonize a space the size of our galaxy, even without faster-than-light travel. (The argument is exponential - we form n colonies, then they form n colonies each, and so on.) Since this is a tiny fraction of the age of the galaxy, there should be evidence of extraterrestrial intelligence everywhere. So, where is everybody?
One could say that perhaps other forms of intelligent life dislike exploration, or that they don't interfere in the affairs of other civilizations, or that they leave no trace of their presence. But the one data point we have (humanity) is keen on exploration, isn't shy at all about interfering in the affairs of others, and is pretty bad at cleaning up after itself.
This is being accepted by a growing number of SETI researchers, who believe (somewhat controversially) that humanity is the most advanced form of life in our own galaxy, at least.
Toronto-area transit rider? Rate your ride.
Perhaps, once a civilization reaches a high enough level of development, its citizens become satisfied and cease interacting with the universe, like an enlightened yogi disappearing into his own navel.
In our case, this is happening with automated production and escapist entertainment. Once we have nanotechnology and perfect virtual reality, we will be able to trick ourselves into eternal happiness, and won't want to bother with anything else.
Any pleasure-driven intelligence which learns to satisfy its survival needs without effort will eventually just turn on its pleasure center and live in perfect contentment.
Not that it really makes a difference. They may be out there, but they don't want to be bothered.
--
The first of these is the Heisenberg Uncertainty Principle. This states that you cannot know both the location AND velocity of a sub atomic particle (protons and on down). Stated otherwise, you cannot measure its location without changing its velocity, and vice versa. Or, you cannot separate the observer from the observed. For example, if you wanted to bounce a photon of light off of a proton so you could "see" it, the photon would have to have an extremely small wavelentgh (smaller than the diameter of the proton). This means that the photon would carry a lot of angular momentum, and when said photon collides with the proton, path of the photon would change, but SO WOULD THE PROTON - and there is no way of knowing in what way that path changed. The upshot of this is that you can never accurately predict what a particle will be doing at a given location. This is why electron orbitals are represented as probability clouds (in actual physics labs) as opposed to the more popular "solar system" model.
What the HUP means as far as processors go is that there has to be a reliable way of controlling (starting, stopping, redirecting, DETECTING) super small particles for it to work. As of right now, even in theory, we have no way of doing that.
Near as I can tell, being a failed physics major, the only way we'll get quantum computers is when we finally crack the GUT's and String Theory (the reconcilliatioin of newtonian and quantum mechanics). Albert Einstein, Steve Weinberg, Richard Feynman, and Steven Hawking have all been stumped by this problem. In fact, Einstein compared it to the biblical "thorn in the chest" because he felt his theories were incomplete while that problem existed.
You know what's even more incredible? More than likely, FTL (faster-than-light) computers are closer. I think that's another post, though.
When nuance becomes the only objective we lose the ability to function
"Even a small probability multiplied by 10^22 gets pretty big."
If only common people understood such things. Be it the chance of two hydrogen atoms fusing in the sun, or the emergence of a technological world...
I remember reading last year that when AMD was beating the pants off of them with the 1G Athlon, Intel was still revved for producing P3-600's, more in line with Moore's law.
Maybe you forget that there isn't a difference in the number of transistors in P3 600's and P3 1000's. Clock speed has nothing to do with Moores "Law".