The Death of the Silicon Computer Chip 150
Stony Stevenson sends a report from the Institute of Physics' Condensed Matter and Material Physics conference, where researchers predicted that the reign of the silicon chip is nearly over. Nanotubes and superconductors are leading candidates for a replacement; they don't mention graphene. "...the conventional silicon chip has no longer than four years left to run... [R]esearchers speculate that the silicon chip will be unable to sustain the same pace of increase in computing power and speed as it has in previous years. Just as Gordon Moore predicted in 2005, physical limitations of the miniaturized electronic devices of today will eventually lead to silicon chips that are saturated with transistors and incapable of holding any more digital information. The challenge now lies in finding alternative components that may pave the way to faster, more powerful computers of the future"
I'll... (Score:5, Insightful)
Re:I'll... (Score:4, Insightful)
Comment removed (Score:5, Insightful)
Re: (Score:3, Insightful)
Re:I'll... (Score:5, Insightful)
Silicone was expensive to refine and manufacture at one point too. Like all new technologies the REAL cost is the in manufacturing and the cost goes down once we've manufactured enough of it to refine the process until we know the cheapest and quickest ways to do it.
Re: (Score:3, Insightful)
Re: (Score:2, Insightful)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re:I'll... (Score:5, Interesting)
Nanotubes have a certain chirality - denoted by (m,n) with m and n being integers. Those two numbers define the properties of the nanotube (e.g. if m-n is a multiple of 3, the nanotube is metallic - otherwise it is semiconducting). They also determine the radius.
So far no one has come up with a way to get a nanotube of a certain chirality. They just synthesize many nanotubes and then pick manually the ones they want - if it exists in the sample. Until they can do this, the nanotube industry will not become a reality.
Re: (Score:2)
Re: (Score:3, Interesting)
Re: (Score:3, Interesting)
Re: (Score:2)
Generally (always?) nanotubes are not grown individually. The synthesis process simply produces many nanotubes. Then you have to pick the one you want among those.
So yes, a nanotube of a certain chirality can continue to grow longer and longer without losing that chirality. But you don't just get one nanotube of desired chirality, but plenty of nanotubes with different chiralities. No one has so far found a way to get just ones of a desired chirality.
Re: (Score:2)
Pet Peeve (Score:2, Informative)
Re: (Score:2)
Unless you're a female robot with a boob job.
Re: (Score:3, Interesting)
What is the problem, you may ask? Well, just look at the Wikipedia entry you linked. Even doped silicon is still %99.999999 pure.
So, you have your gate that is thousands of atoms in volume, and dopant concentration that is in the 1 million to 1 billion ratio...so what is the likelyhood that your gate is going to contain tha
Re: (Score:2)
Re: (Score:2, Insightful)
Re: (Score:2)
Also, crystal silicon was first obtained in 1854.
Re:I'll... (Score:4, Informative)
No one said anything about mass production in 1916, read the post again.
We started learning to purify it in the 1910's.
From Wikipedia:
The earliest method of silicon purification, first described in 1919 and used on a limited basis to make radar components during World War II, involved crushing metallurgical grade silicon and then partially dissolving the silicon powder in an acid. When crushed, the silicon cracked so that the weaker impurity-rich regions were on the outside of the resulting grains of silicon. As a result, the impurity-rich silicon was the first to be dissolved when treated with acid, leaving behind a more pure product.
From: http://en.wikipedia.org/wiki/Integrated_circuit [wikipedia.org]
The first integrated circuits were manufactured independently by two scientists: Jack Kilby of Texas Instruments filed a patent for a "Solid Circuit" made of germanium on February 6, 1959. Kilby received several US patents.[4][5][6] Robert Noyce of Fairchild Semiconductor was awarded a patent for a more complex "unitary circuit" made of Silicon on April 25, 1961. (See the Chip that Jack built for more information.)
The first integrated circuits contained only a few transistors. Called "Small-Scale Integration" (SSI), they used circuits containing transistors numbering in the tens.
SSI circuits were crucial to early aerospace projects, and vice-versa. Both the Minuteman missile and Apollo program needed lightweight digital computers for their inertial guidance systems; the Apollo guidance computer led and motivated the integrated-circuit technology, while the Minuteman missile forced it into mass-production. These programs purchased almost all of the available integrated circuits from 1960 through 1963, and almost alone provided the demand that funded the production improvements to get the production costs from $1000/circuit (in 1960 dollars) to merely $25/circuit (in 1963 dollars).[citation needed] They began to appear in consumer products at the turn of the decade, a typical application being FM inter-carrier sound processing in television receivers.
The next step in the development of integrated circuits, taken in the late 1960s, introduced devices which contained hundreds of transistors on each chip, called "Medium-Scale Integration" (MSI).
They were attractive economically because while they cost little more to produce than SSI devices, they allowed more complex systems to be produced using smaller circuit boards, less assembly work (because of fewer separate components), and a number of other advantages.
Further development, driven by the same economic factors, led to "Large-Scale Integration" (LSI) in the mid 1970s, with tens of thousands of transistors per chip.
Integrated circuits such as 1K-bit RAMs, calculator chips, and the first microprocessors, that began to be manufactured in moderate quantities in the early 1970s, had under 4000 transistors. True LSI circuits, approaching 10000 transistors, began to be produced around 1974, for computer main memories and second-generation microprocessors.
Re:I'll... (Score:5, Funny)
Re:I'll... (Score:5, Insightful)
Yes, and we're so damned good at manipulating it. All this newfangled stuff is pie-in-the-sky at this point. Yes, I suppose we'll eventually replace it for the likes of high-end processors, as you say, but everything else out of silicon for a long time to come.
People keep bring up Moore's Law, as if it's some immutable law of physics. The reality is that we've invested trillions of {insert favorite monetary unit here} in silicon-based tech. Each new generation of high-speed silicon costs more, so that's a lot of inertia. Furthermore, if Guilder's Rule holds true in this case (and I see no reason why it shouldn't) any technology that comes long to replace silicon will have to be substantially better. Otherwise, the costs of switching won't make it economically viable.
Re:I'll... (Score:4, Informative)
For example:
If it costs 1/100 the price but seen no end users gains in 'speed' and/or 'power' it could replace silicone. It's not better at doing anything, it just has a higher value.
"All this newfangled stuff is pie-in-the-sky at this point."
hmmm, some of this is a lot farther along then pie in the sky.
Most people on
Finally:
The real problem with silicone is the fabs. They are running into some serious problems at these incredibly small sizes. Some fabs are problems with metal atoms in the air, atom that are below detection and the ability to remove.
I am not dooming and glooming silicone here(although there are some advantages to hitting a minimum size) it's just that some problems aren't going away and are getting harder to deal with and the past work a rounds aren't cutting it.
Re:I'll... (Score:5, Interesting)
Re: (Score:2, Insightful)
absolutely positively undeniably 100% wrong
Just because your garage door opener can't "solve" Folding@Home doesn't mean that we can't dream. I mean, at some point we truly need to be able to say something like "well my garage door opener has more processing power than BlueGene/L did in 2008"
Seriously, get over yourself and your "reality"
Re: (Score:3, Funny)
Re: (Score:3, Funny)
sometime in the 1Thz-Garage-Door-Opener-Overlord-future:
GARAGE_OWNER: "Open the garage door please, Hal"
GARAGE_DOOR: "I'm sorry Dave, I can't do that.
You're saying that you want this sort of thing to happen? No thanks. I like my appliances simple and mute, thankyouverymuch.
Re: (Score:2)
"Open garage door, please, HAL."
"I'm sorry, Dave, I can't do that."
(pause)
"Why not?"
"I think you know the answer to that question."
Re: (Score:2)
Guess people's sarcasm detectors aren't working.
Re:I'll... (Score:5, Insightful)
The real advantage of silicon for many years was that SiO2 was/is a decent gate materal for mosfets and insulator for insulating the metal from the main body of the IC and could be grown easilly on the surface of silicon. But afaict this advantage has dwindled as we need CVD deposited insulators for insulating between multiple metal layers anyway and as processes have got smaller there is a push to switch to other gate materials for better performance.
The main advantage of silicon right now is probablly just that we are very used to it and know what does and doesn't work with it. Other semiconductors are more of an unknown.
Even if silicon gets displaced from things like the desktop/server CPU market though I suspect it will stick arround in lower performance chips.
Re: (Score:2)
1: obviously a bigger chip costs more to make, you need more semiconductor grade silicon, more space on the masks, more space in the furnaces etc.
2: a larger chip has a higher possibility of being struck by a defect increasing the reject rate and pushing up costs still further.
Making the wafers bigger helps will some of problem 1 and indeed wafers have to bigger over the years. But not all of it and brings problems of it's own (the larger a wafer is the more cha
Re: (Score:2)
I don't know what the velocity of signals in the tracks on a silicon chip are but I would expect it is somewhere arround 10^8 m/s. It will definitely be lower than 3x10^8 m/s
10^8 m/s is about 10mm/ns . When you consider that modern processor clocks have a cycle time arround 0.3ns and that modern processors often already have fairly big die
Re: (Score:2)
Re:I'll... (Score:5, Interesting)
Agreed. Besides, they've been saying this since the 90s, that silicon can't possibly get any faster and it'll be replaced very soon.
I call BS. They had 350 gigahertz silicon chips 2 years ago [news.com]:
"At room temperature, the IBM-Georgia Tech chip operates at 350GHz, or 350 billion cycles per second. That's far faster than standard PC processors today, which range from 3.8GHz to 1.8GHz. But SiGe chips can gain additional performance in colder temperatures....SiGe chips, the scientists theorized, could eventually hit 1 terahertz, or 1 trillion cycles a second."
I think silicon is safe for awhile longer.
Re: (Score:2, Interesting)
Re: (Score:2)
However, the record for fastest transistor has been held by III-V based transistors (i.e. not silicon) for a few years now. See this [sciencedaily.com], for example.
So the article's not all that wrong.
Re: (Score:3, Interesting)
To give you an idea, in a mixed signal BiCMOS chip where the digital components are standard CMOS and there's a SiGe layer on top for the RF circuits, the RF transistors are capable of amplifying an input sine wave all the way to the multiple tens of GHz. In the same proce
Silicon most common? No. (Score:2, Funny)
Re:I'll... (Score:4, Funny)
There, I fixed that for you.
Re: (Score:3, Funny)
Re: (Score:2)
Wait a minute, the monitor I'm staring at is a vaccuum tube. They told me vaccuum tubes were gone a couple of decades ago and they're still in guitar amps, too.
I predict that this prediction about the demise of silicon is as accurate as their predictions about the demise of vaccuum tubes. But in four years nobody's going to remember their prediction, or mine either.
Re: (Score:2)
Wow grandpa, your face is so tan!
Re: (Score:2, Insightful)
Re: (Score:2)
Re: (Score:2)
Believe it when you see it... (Score:2)
That's kinda of like saying the sequel to Duke Nukem Forever is going to be the best game ever.
Let them speculate ... (Score:5, Insightful)
In the meantime, other researchers will figure out ways to make silicon work smarter, not harder.
Re:Let them speculate ... (Score:4, Funny)
Re: (Score:2)
It's hype, nothing but.
I'd like to see something that is vastly better, cheaper, more energy efficient, and capable of greater performance... but until that c
Re: (Score:2)
So if we want better chips, we have to continue to cram in those transistors. Since we're coming up against the 1
Re: (Score:2)
I tend to agree. Amdahl tried to develop multilayer chips (essentially a three dimensional layer-cake approach) but failed, for a variety of reasons, although I remember reading that it was a complexity issue (inadequate design tools) as much as failure rate.
That's probably a logical way to continue increasing complexity: just stack extra circuitry vertically. Discards don't matter much if you have
Re: (Score:2)
Re: (Score:2)
Sorry, what I meant was having the software tools to develop applications to make effective use of all that power. We're having a hard time writing compilers that can parallelize code across eight or nine processors: what if we have chips with a thousand cores? Of course, with that much power it probably won't matter much if your code isn't all that efficient.
Re: (Score:2)
For the past thirty years I've been hearing about the "ultimate limits" of silicon-based technology, and the various advancements using {insert favorite exotic material here} that were going to supplant it. I have some history on my side when I say that predictions of silicon's demise have always proven premature, and I'm not convinced that this time is any different. If I'm wrong, great
Didn't slashdot already cover it? (Score:2)
Not again (Score:5, Informative)
And of course what's really reaching a limit is not the CPU's, but our ability to use them effectively. See "TRIPS architecture" on the wiki as an example end-run around the problem that offers hundred-times improvements using existing fabs.
Maury
Re: (Score:3, Insightful)
Re: (Score:2)
So this other technology that claims it will be going by 2012, it's not going to happen, but this other technology that claims it will be giong by 2012 is a show in!
Sorry, you will nede more then that. All these slashdot articles remind me of when tubes went away*, the same arguments.
*Yes, I KNOW there are devices that use tubes, seriously.When was the last time you saw a tube tester in a grocery store?
Re: (Score:2)
gaas - a little nostalgia. (Score:3, Informative)
Re: (Score:2)
As I understand it (I'm no expert; I could easily be wrong) GaAs starts to lose its speed advantage below the 0.35 micron node because the drift velocity saturates.
Re: (Score:2)
Re: (Score:2)
For sure... after many decades we _are_ approaching real physical limits. But I guess for me the real question is "so what?" Or to put it another way, "was silicon ever the problem?" My own computer is way faster than the rest of the computer it's attached to, and the bottlenecks are almost always either HD or GPU related.
If we really can extract vastly more performance though architecture changes, then at some point you have to do a price/perfo
Next phase: Transistors in Silicone (Score:1)
Overclocking might be fun as well: "Hey, I managed a stable DD at room temperature!"
Re: (Score:2)
Overclocking might be fun as well: "Hey, I managed a stable DD at room temperature!"
In that context, you probably should have said, "overcocking". I know I get enough emails on that subject every day.
Again? (Score:1)
(Okay, dating myself here, but still...)
Much sillio articulo (Score:5, Insightful)
Unlikely (Score:5, Informative)
Intel's CTO Justin Rattner just gave a talk at Cornell two days ago; he covered this topic carefully and confirmed that Intel has the technology and plans to carry out Moore's Law for another 10 years on silicon. Technologies such as SOI [wikipedia.org] and optical interconnects will be leveraged to hit this.
It's not necessarily the size of the transistors that make chips hard to make these days either (although they are now giving us huge problems with leakage current). It's harder to route the metal between these transistors than it is to pack them onto the silicon. New processors from Intel and AMD have areas with low transistor density just because it was impossible to route the large metal interconnects between them. Before we can take advantage of even smaller transistors we'll need a way for higher interconnect density.
Re:Unlikely (Score:5, Interesting)
It would be a stock hit to say "We will be replacing silcone in x period of time if X is any longer then 'right now'.
Some new technologies solve those problems. Technologies in the 'we hobbled something together proof of concept stage, not the I wrote this down on paper stage.
Some of it is impressive, whether or not there will b a practical way to mass produce it is another thing. If not, I can imagine a time in the future where only large entities that can afford 500K a chip will be using them. Or anyone at home that can afford the latest electron microscope, laser, super cooling.
meh, I'm just glad the MHz war is pretty much subsided and we are FINALLY focusing on multi-core.
Not every chip needs speed (Score:3, Insightful)
Re: (Score:2)
And the rest of the industry doesn't need extreme speeds - there are microcontrollers, integrated buffers, logic gates, comparators, operational amplifiers and loads of other $0.05 crap you got in your toaster oven, blender, wirst watch, remote-controlled toy car, printer, Hi-Fi, etc., etc. And there is an obvious priority for those: cheap and reliable. So the silicon is not going anywhere.
And let's not forget Solar Cells, which are increasing production like crazy (and is causing silicon prices to increase).
Re: (Score:2, Insightful)
Birth vs. Death (Score:4, Insightful)
Wrong tag (Score:3, Funny)
While we're at it, might add that Duke Bend'Em Forever tag, too...
Yawn (Score:2)
ECHO! Echo! echo! (Score:5, Insightful)
This has been getting bandied about every time someone comes up with a new, spiff-tastic technology/material to build an IC out of.
"THIS COULD REPLACE SILICON! WOOT!"
Yet it keeps NOT happening. Again, and again (and again).
The trailblazers keep forgetting, the silicon infrastructure has a LOT more money to play with than a given exotic materials research project. And, in many cases, what's being worked on in exotics can be at least partially translated back to silicon, yielding further improvements that keep silicon ahead of the curve in the price/performance ratio. Additionally, we keep getting better at manufacturing exotic forms of silicon too.
So, until silicon comes to a real deal-breaker problem that nobody can work their way around, I SERIOUSLY doubt that silicon IC is going anywhere. Especially not for a technology that has taken several years, and recockulous amounts of money simply to get a single flawless chip in a lab.
Not so fast... (Score:3, Insightful)
For something else to replace silicon it will have to not only be better, but so much better that it will justify the investment, or it will have to offer other, significant benefits, like being cheaper to produce, using less power or being smaller. Of these, I think speed is probably the least important, at least for common consumers.
Personally, I still haven't reached the point where my 3 year-old machine is too small or slow - not even near. It wouldn't make sense to upgrade, simply. I think most people see it that way, they would probably be more interested in gadgets than in a near-super computer.
Re: (Score:2)
Well, now that depends. If you mean a supercomputer whose only function is to run Microsoft Office faster
Four years, eh? Then what? (Score:3, Insightful)
When someone makes a nanotube 80486 that I can buy and use, THEN I'll start to believe we're close to a technology shift. Hell, give me a 4004 - at least it's a product.
Bottom line: We're not there yet.
to put things in perspective ... (Score:2)
solar panel production should benifit (Score:2)
I wonder if those same silicon wafer production facilities can be converted to make solar panels once the move away from silicon in the microprocessor industry takes place?
If True (Score:2)
Silicon Scaling (Score:2, Interesting)
Intel thinks we may hit the limit by 2021. http://news.zdnet.com/2100-9584_22-5112061.html [zdnet.com]
Re: (Score:2)
Homebrew graphene transistors (Score:3, Interesting)
Re: (Score:2)
The personal automobile is dead (Score:5, Informative)
Re: (Score:3, Funny)
Only half of the story. (Score:2, Insightful)
In addition to some of the points made by other posters (Silicon CPU's will live on in smart systems, cheap systems, handheld systems, etc.), there is a whole world of silicon chips that are *not* CPU's! Analog and mixed signal circuits need highly linear devices--not just switches
my Germanium stockpile (Score:2)
Re: (Score:2)
Huge fallacy - Moore's law strikes again (Score:2)
This statement:
...the conventional silicon chip has no longer than four years left to run... [R]esearchers speculate that the silicon chip will be unable to sustain the same pace of increase in computing power and speed as it has in previous years.
Does not equal this statement:
Hardware: The Death of the Silicon Computer Chip
What the first statement means is that they may have found something faster than silicon chips. That doesn't mean that silicon will suddenly "go away" just because it cannot maintain Moore's law predictions.
Hell - do you think they're going to put some uber carbon nanotube processor in your TV remote or your microwave oven control panel? Silicon cpu chips have *plenty* of uses other than high end mainframes. They're damn useful - that's why they're
Silicon is far from exhausted. (Score:2)
For starters we're still using a primarily two-dimensional structure on the surface. There's no reason you couldn't build your structures in three dimensions through a cube of the material. (Yes power and cooling become issues, but they're soluble.)
Going truly 3-D again shortens the wiring, leading to another speed increase with a given speed of components.
could you make me a billion for a $100? (Score:2)