Slashdot Mirror
Intel Researchers See Moore's Law Becoming Obsolete
prostoalex writes "A paper, published by Intel researchers, claims we might be the witnesses of Moore's Law becoming obsolete, as the rate of shrinkage for transistors goes lower with each year. In 2018 we might be able to get the chips manufactured with 16-nanometer technology, then one or two more manufacturing processes will shrink it even further, but after that we're facing the physical limits."
We keep hearing this over and over again, and yet there's always a new technological breakthrough that lets the trend continue. This is talking about 2018...Quantum computers anyone??
We may be getting smaller, but as this happens we'll need higher voltages to force things to happen on that level. And with those increased voltages (and the problems of things being crammed so tightly together) we'll see the effects of those electrons in such close proximity resulting in errors. Sure, maybe we won't hit a brick wall for a while as far as how much we can cram onto a chip, but what about the logistics? Will it really be worth the effort if we can't rely on these little marvels to remain accurate?
Damon,
http://actionPlant.com
... that drives people to try to pinpoint the exact coming moment when it will become obsolete? I suspect it is a desired to tack their own 15 minutes of fame to the long-lasting fame Moore has enjoyed.
I was taking one day at a time, but then several days got together and ambushed me. (from a Rhymes with Orange comic)
because if Moore's law continued forever, it would prove P=NP. Think about it.
Even if there were no way to manufacture chips smaller/faster than the ones we have today, there are always going to be refinements in the manufacturing process, making chips cheaper and cheaper. There are always supercomputers. Perhaps, also, we could find a way to really minimize waste heat, allowing many CPUs per board.
It's also possible that DNA computation and other kinds of biocomputing are going to come along. These have the advantage of being gigantically parallel; they would possibly be good for tasks that are not latency sensitive but require immense brute force.
I'm satisfied that we have enough axes of advance to keep progress moving forward. Remember, computers have only been around for a very short while; I refuse to believe that we hit on the fitness maximum on the first try; there have to be technologies out there that are far faster/cooler/smaller.
Intolerance for ambiguity is the mark of the authoritarian personality.
Once we approach the phyisical limits, we can simply expand in a different way. Just start adding CPU cores to the machine. SMP boxes are becoming fairly common already, even the in the PC market, and I definatly see that trend continuing. Once things get cheap enough, why not stick 16 or 32 chips in a machine? Heat and power issues can be minimized by greatly UNDERclocking the chips. In another few years, chips will be at insane frequecys, and instead of pushing them the limit by running that at super high power levels, just back things off a bit.
TODO: Something witty here...
I remember sitting in a lecture in 1997, where some luminary from IBM predicted the death of Moore's Law in 10 years. Now it's 2003 and the death of Moore's Law is being predicted in 15 years.
Technologically, there will probably be enough clever ideas to take chip manufacturing beyond the point where it is no longer economical to make such fast processors. Consider that in 1980, a handful of engineers could sit down with pencil and paper and design a microprocessor. Today it takes teams of PhDs in physics, math, and engineering to do the same, in multi-billion-dollar facilities with the latest design tools and techniques. One day the buying public will realise that e-mail and word processing does not need a bazillion gigahertz, and gamers will have photorealistic animation with excellent AI. The chip maker will not make back the investment on a fab plant, and on that day Moore's Law will be dead, not for physical reasons but for economical ones.
Toronto-area transit rider? Rate your ride.
Silicon is dead. Long live diamonds!
Of course, because diamonds are forever!
I find it interested that just because Intel thinks it has reached the limits of its ingenuity that Moores law will become obsolete. As you say, if they don't do it, some other company will. Especially since they have so much money tied up in silicon, another competitor with less capital tied up could emerge.
- electomechanical calculators
- relay based computers
- vacuum tubes
- discrete transistors
- integrated circuits
Moore's law will continue, but it will continue based upon a new paradigm that sweeps in and seems to "miraculously" preserve Moore's law. The obvious next step is three dimensional integrated circuits and there is already research in exactly that direction: Intel's 3d gates. AMD is also in the game. When 3d transistors lose steam some new paradigm will take its place.Moore's actual Law does not require ever-shrinking transistors. It only requires that we put more of them into each chip. Double-sides chips, multi-die packaging, or 3-D layering of circuits would help increase the number of transistors in each "chip." You may think that multi-die chips is a cheat, but when it comes to packing in several billion transistors into a CPU, who cares how they do it.
Two wrongs don't make a right, but three lefts do.
Most of you know this, so please just bear with the sermon for those who do not.
Moore's Law is a marketing term which was coined by the press, not Gordon Moore himself. It's not a law in the scientific sense, like the Law of Gravity. The 'law' simply states that the number of transistors on IC's roughly doubles every 18 months. People have been predicting the death of Moore's Law for many years, and probably will for many more.
If it truly were a law, it could not die. But eventually it will fail. In the mean time, it's a 'law' that keeps sales and marketing people busy, ensuring there will always be faster processors to run the latest bloatware.
Ruby on Rails Screencast
Intel is becoming obsolete. Intel's steadfast opposition to changing their (unbelievably ancient) chip architecture and/or changing their manufacturing processes radically enough to actually innovate is no reason to declare the imminent failure of their competitors.
This is how you visualize an electron tunnelling across a gate:
Heisenberg's uncertainty principle says that we can't know an electron's position accurately. There's always a little bit of uncertainty about where it is. So, imagine the position of an electron not as a point, but as a little 'O'. That circle is the area that the electron could be. At any time it could be in any random place in that circle.
Now, if the 'O' is centered on the edge of one side of the gap, and the gap is bigger than the circle radius, then the electron has zero probability of crossing the gap. But, once the gap is smaller than the radius of the circle, then you've got parts of both sides of the gate within the area of the circle. Since the electron can appear randomly anywhere inside the circle, that means that sometime that electron will appear on the other side of the gate. As the gates get smaller, the probability that the electron will randomly appear on the other side of the gate goes up, until so many electrons are crossing the gate that we can't tell if the thing is on or off.
This is America, damnit. Speak Spanish!
4 and 10 Ghz is a huge jump. I doubt Intel would release them that close together. It would be horrible marketing sense. Why make such a big bang jump to 4 and 10 when Intel can suck much more money producing a 4 Ghz then a 5 Ghz and the 6 and so on. Indeed I am questioning your source, but time will tell if you are correct about these releases. As far as Moore's law: In the past when people have said Moore's law must stop it was because researchers were having harder and harder times finding ways to product smaller chips. Now we are getting close to the point that we are arranging the silicon semiconductors atom by atom. Once your organizing atoms you physically cannot do much more. You cannot work with smaller components than on an atom by atom basis. Researchers have trouble even isolating the constituent parts of an atom, and the components of an atom are still highly theoretical. And those components that have been identified are highly unstable. Supposedly though there is something called quantum computing. I don't understand it but maybe quantum computing which doesn't use transistors (as far as I know) will be the future.
So you're saying that exponential growth can be sustained forever?
"They redundantly repeated themselves over and over again incessantly without end ad infinitum" -- ibid.
Your point is exactly what makes the necessity of the "next big thing" essential to the survival of that industry, and which leads me to believe that we'll see the computer industry wind up look a lot like the auto industry. We saw great advancements in the first several decades, but nothing that really changed how the core machine worked; we simply spent 100+ years refining it and improving efficiency and power (and safety). Of course there were always those exotic electic cars, but their use never become too widespread.
Now we're finally on the verge of the next big step; fuel cell autos. Just like they expected cars to fly fifty years after they were invented (but with no real change in the actual technology of the machine), so now we're expecting exotic things like quantum mechanics to be commonplace in computing environments in twenty years.
I think rather we'll see companies settle in; the big ones will survive if they're smart, while others will come on the market with their own claim to fame; shapes, colors, "safety" ratings, and finally government efficienty mandates. It could well be 100 years of "getting it right" before we finally see widespread implementation of a completely new technology.
Damon,
http://actionPlant.com
Actually what the guy said is close enough to the truth. It explains the general idea without getting to details, and favors readability over correctness. I bet that 10 years ago a somewhat similar explanation was given to you, and it provided what was meant to provide: a first step.
When a scientist says that something is possible, he is most probably right. When he says that something is impossible, he is probably wrong.
- Arthur C. Clarke
While I think that quantum tunneling effect is likely to place limits on the size of electronic gates, who says we have to use electronic gates?
(BTW, why are we worried about AI when our I is suspect in the first place?)
Because *we* are the "I" attempting to create the "AI". That worries me.
It is also inevitable that advances will never actually stop, but this article is all about the rate, which so far has been exponential. And any exponential growth is not long-term sustainable.
It's a funny coincidence that Moore's Law will hit the wall (S-curve actually) at about the same time that nanotechnology is maturing, allowing for the next paradigm in computing to continue our exponential progress.
Molecular manufacturing -- while still 10 to 20 years away -- means that billion-dollar factories won't be needed to manufacture ANYTHING anymore. Everything, from food to clothing to genetically evolved open source 3D chip designs, will be built bottom-up for the same lowcost as growing a potatoe.
--
Power to the Peaceful
>either way, moore's law is dead. Kurzweil suggests that after moore's law, an exponential boost will occour, with the advent of a new technology, not simply in materials, new research all together, in accordance with the aforementioned "thillbert's law" ;)
:)
So, in other words, the Moore's Law that Moore actually stated (about shrinking transistors) is about to become false, but the Moore's Law that people think of when the phrase is spoken (about increasing power at a constant price) is likely to continue into the indefinite future.
The point is not to extend the time it takes to reach the 5nm limit, beyond which no material will allow further shrinkage. If we don't reach that limit as fast as Moore's Law predicts -- if it takes several more decades to reach 5nm as you suggest -- then Moore's Law will have already failed.
In other words, Moore's Law says that progress will occur at a certain (very fast) rate, not just that progress will occur. If you take longer to make progress than Moore's Law predicts it should take, then Moore's Law has failed.
"Those who have never entered upon scientific pursuits know not a tithe of the poetry by which they are surrounded."