Slashdot Mirror
A Well-Chilled 750GHz Feasible Within 5 Years
drkhong writes: "...at least if you've got a good cooling system. IEEE Spectrum has an article about the next generation ICs. Using superconducting materials cooled down to 5K a peak of 750GHz has already been reached. Just think about how far light goes within one clock cycle, and then tell me you aren't impressed." These low-temperature devices are made of niobium
(a superconducting metal), and use something called Josephson junction devices, resulting in chips for which the article states "there are no known physical barriers to decreasing size by a factor of 10 and thus increasing speed by a factor of 10, using lithography to move from today's 3-m linewidth to 0.3 m."
I mean it's cool and all... but I really don't want a desktop that can kill me if it has a coolant leak? It'd be great for supercomputers tho' =)
.technomancer
This is, of course, very impressive but let us not forget that this requires cooling down to five degrees Kelvin. We are well past heatsinks and fans at this point. Unless the prices come down, it will cost around twenty THOUSAND dollars to cool the chip down this much.
It will be a long time before you see a system like this on your desktop. Unless we develop room-temperature superconductors, of course. But that would change everything...
Oceania has always been at war with Eastasia.
The article talks about a new Hypres (sp) AD converter that runs at 12 GS/s, and can dynamically change SNR for bandwidth, and vice versa (or at least # of bits, as the article says).
Do you have any idea how many bits this puppy can do per sample? I didn't find this number, strangely, within the article.
Thanks.
make world, not war
Applying concepts like that to the brain is useless. The brain is unbelieveably parallel, and runs asynchronously, so "how fast can we think?" really doesn't mean anything. I know you were just thinking of the sci-fi aspects of this and all, but this is one nut that just ain't gonna be cracked.
Any predictions about RAM or Hard Drives in five years? Compared the progression of CPU speeds, increases in memory speed and capacity has been almost static. Perhaps RAMBUS can start suing for royalties even before they develop their PC750000 RIMM
Nope. That's the whole point of 128 bits. Assuming you're doing brute force, and have 1000 of these computers overclocked to run at 1000 GHz (to make the math easier), and of course assuming that they can do one trial per cycle,
2 ^ 128 trials * 1 cycle / trial * 1 second / (10^12 aggregate computer cycles) * 1 year / (3600 * 24 * 365) = 10790283070806014 years = 10 quadrillion years.
There are methods that the optics community uses to measure high speeds / short times. One such idea is an autocorrelator, which interferes a fast signal with a delayed version of itself, thus allowing you to map out the waveform shape. Possibly the 750GHz team made a version of one of these.
There are pulsed lasers that have pulse widths of less than a femtosecond (admittedly, you probably have to build these yourself and be careful with components). these are probably the best generators of delta functions, and are also probably used to probe some of these novel devices.
make world, not war
I submitted a more accurate review that, among other things, didn't confuse data rate and clock frequency, to Slashdot and Kuro5hin last week. Kuro5hin accepted it. You can read it and reader responses at:
/ 12/10/0925/1544
http://www.kuro5hin.org/?op=displaystory&sid=2000
Yea! Since I just saw a factoid saying Windows 2000 is the largest 'program' ever writen (I'd say it's multiple programs...) - from Learning Kingdom - "A complete printout of its 29 million lines of source code would form a stack of pages 193 feet high (59 meters), about as tall as a 19-story building." So are faster processors a good thing? That is, doing a whole bunch of dumb things really fast is not nesc. better than doing a few smart things slower. Tell me about how x86 is being utterly thrown out the window and that will be exciting.
closed minded is as closed minded does
Well, now that we have macroscopic quantization of magnetic flux, and fluxon-switching devices, that Flux Capacitor can't be far off now can it? Anyone got a DeLorean sitting in the barn waiting to be refitted? Now's your chance!
--frank[at]unternet.org
When using light primarily instead of electricity, a lot less heat is generated. Would some of the more advanced chips be cool enough to run safely without a heat sink? If so, this would not only extend the life the standard CPU but slightly reduce the ammount of clutter in the standard PC box. //this message posted by powerpenguin from somewhere.
-----------------------------------------
Perversely greped and groped by PowerPenguin
There are two huge problems with superconducting logic that don't seem solvable in the near future. They are:
1. Cost : These things are enormously expensive to manufacture and operate, and it is the economy of scale of CMOS techology which has enabled, more than anything, the current computing revolution. Do you have any idea how expensive coolant and the dewar to use it in are to get something to 5K? Even the so-called "high-temperature" superconductors have to be pretty damn cold to function; they just don't need to go so close to absolute-zero.
2. Integration This is probably the killer. It will be extremely difficult to integrate many devices together. Even if myriad technical difficulties are overcome, the solution is not likely to be inexpensive, as CMOS technology is. For III-V semiconductors (which use much less exotic materials than superconductors), high defect rate, problems with lattic matching of the materials, and the lack of a high quality native oxide (like SiO2 in silicon) have made it impossible to achieve integration levels anywhere close to that achieved in silicon. Even GaAs, the most well-understood III-V semiconductor, can't be integrated to more than a few thousand devices. That's why we don't have 20 GHz GaAs microprosessors. And superconductors are even HARDER to deal with.
In summary, even if researchers are able to overcome almost insurrmountable odds to find away to reliably integrate meaningful numbers of these devices on a single die, I think it is very unlikely they will be able to do it cheaply, which is just as important as being able to do it at all. Otherwise, this technology will be of interest only to the military.
By the way, I know III-V semiconductors have a lot of very important uses, especially in optics and RF. It is a fact, however, that III-V logic is mainly of interest to the military and the space industry.
Even that's not correct. Electrons in copper move very slowly, on the order of centimeters/minute. Ripples, on the other hand travel at approximately half c (or thereabouts).
As another user already pointed out, even if this thing were one million times faster than Intel and AMD's current stuff (which it isn't), breaking 128 bit encryption would still take time well beyond the death of the universe.
Doesn't this wildly violate it? Right now we're at 1.5 GHz. Moore's law states that this will double every 18 months. That's about three times it should double in the next five years. La la la, David does some simple math:
12GHz.
Hmm. I know that Moore's law is just a rough estimate, but so far we've stuck to it pretty faithfully, right? If the rate of increase is increasing (aaaahhhhh! semantically difficult sentence!), I'm gonna be really impressed with where our technology goes!
Unless they mis-estimated their release date (and we know that's never happened) by about six years.
It may look like I'm doing nothing, but I'm actively waiting for my problems to go away.
--Scott Adams
Close, but actually the real market for 750Ghz in the home is Quake players who think polygons are 'not sufficient' and want real-time ray tracing at 60fps. I'm one of them. Is 750Ghz going to be sufficent for 'toy story' level rendering in real time?
A pizza of radius z and thickness a has a volume of pi z z a
I have to go to the drugstore to buy a few more pounds of liquid helium, I'll be back after lunch.
Hmmmm... then does that mean that if you never return you'll be missing, presumed fed?
Ceci n'est pas une sig.
I suspect that advances in insulating materials, most of which have been recently declassified, will make cryogenic circuitry usable in smaller, possibly even desktop-sized machinery.
Consider that the entire integrated circuit set could be built onto a substrate (standard thick-film assembly modded for low temperature) and then the entire package surrounded with aerogel or foamed silica insulation. A heat pump or chiller (Peltier devices, sterling cycle fridge, whatever) is attached through a window, and then the heat removed from the device. Given the kind of heat transfer rates you can get now, the heat pump section would only have to be slightly larger than the max power consumption of the circuitry.
Okay, so it will take ten minutes to cool down to the point where the main processor works. This would be any different from waiting for Windoze to boot?
*whup* "Get along, little electrons. Heeyah!"
Well, several things happened:
- Nobody figured out how to make reseasonalbe P-channel devices.
- Small geometries were much harder, because III/V and II/VI type (more than one element) semiconductors suffer from a whole bunch of problems where, which in my limited understanding (I'm a circuits guy) are due to the wrong atom at a place in the crystal lattice, such as a Ga where and As should have been.
- It's easy to take Silicon Dioxide (glass) for granted, until you try to figure out a good way to make insulators on other materials.
- All the while, good ole Si kept getting better and better... not only faster, but higher densities. Today's CPU speed is as much a function of using lots of transistors as it is their speed. As more transistors were available, everyone invested a lot of research and thought into ways to use them to run code faster (superscaler architecture, branch prediction, out-of-order and speculative execution, etc)
Now I've been watching the J-junction for several years now, though I know much less about how it really works that I ought to. I do know there's a big difference between a test device and processes that produce only thousands of them to being viable for a modern microprocessor. GaAs transistors are hugely popular for RF applications, where you only need a small number of them. Today nobody believes the world will eventually be overtaken by GaAs based microprocessors.It seems unlike the world will really be overtaken by J-Junction microprocessors, at least in our lifetimes. Maybe that's just wishful thinking, since I've got a lot of energy invested in transistors, and with a bit of luck that'll remain valuable for another 25 years... but then again, look what happened to all those guys how only knew about tubes!
Anyways, the point is that there's a big difference between a small number of insanely fast test devices to a high density processor with all the other requisites to make a reasonable microprocessor.
PJRC: Electronic Projects, 8051 Microcontroller Tools
use a 750ghz bewolfe to heat your office building!
A few years ago in Minneapolis a company (Honeywell?) decided to shut down their old mainframe because they discovered they only had one job still running on it, eaisally ported to new machines. The day before the final shut off the janitors discovered that the building was built without heaters because the comptuer gave off enough heat to need cooling even in the coldest Minnesota winter. They ended up selling time on the mainframe (for peanuts, not even recovering energy costs) for 2 more years until they could install a heater.
"Honey, We're going to have to take out the kids room in order to put in the new 300 gallon liquid N2 dewar I just ordered... What?"
Eric
Make it idiot-proof and someone will build a better idiot.
The Freon!
This is absurd. The cost of keeping such a superconductor at 5 K is going to keep the general public, and even most corporations, from buying this technology. It's expensive to keep a box at that temperature in the lab (I should know, I'm a chemist). Only the US government would be willing to shell out the money for these low-maintenence devices (maybe). Corporations would rather just use the money to buy the computing speed in multiple CPUs rather than as one - it'd be a hellofa lot cheaper.
Perhaps the people working on the project will eventually be able to use a superconducting material that works at liquid nitrogen temp instead of niobium (perhaps a yttrium complex like we use now? - I don't know the specifics of this 700GHz IC or whether it would be able to use Yttrium complexes). In that case, the cost will go down and perhaps we'll see more corporations buying this tech. In order for personal consumers to buy a 700GHz computer, we'd have to have room-temp or near-room-temp superconductors.
But then we run into one of the hugest physics problems of the late twentieth century. The scientific community no longer has the enthusiasm it once had for searching for that "perfect" superconductor.
---
---
I'm just an ordinary man with nothing to lose.
"Honey, with this new cryogenics tower computer, I'll be able to get my computing done faster, so I can then spend more time on YOUR needs."
:^)
Simple.
At those speeds, interconnect latency is a major problem. A photon can only travel 40 microns during a single clock cycle of that puppy. So chip layout becomes extremely important.
General Relativity: Space-time tells matter where to go; Matter tells space-time what shape to be.
Dec SciAm (not online) shows nanotaubes a 1/20th
the diameter of current wiring and very fast.
There's an additional factor of 2 because, on average, you only need to check half the possibilities. So it's actually 5 quadrillion years. See? Now that seems much more achievable!
Given one hour to live, the student replied: "I'd spend it with professor FP who can make an hour seem like a lifetime."
They used Fluorinert. The -100C overclockers were playing with that.
I also think one of the Crays used artificial blood plasma as the coolant.
General Relativity: Space-time tells matter where to go; Matter tells space-time what shape to be.
--
A feeling of having made the same mistake before: Deja Foobar
With the advent of the 300MHz processor, the 233 I purchased became dirt cheap. Now that there are 1.5 GHz chips out, you can get an 800 MHz chip dirt cheap. When the 750GHz chips are produced, I will be lined up to buy an obsolete 500GHz chip that will be fast enough to start windows from boot in less than three minutes! Yay bleeding edge subsidizing second-stringers!
It may look like I'm doing nothing, but I'm actively waiting for my problems to go away.
--Scott Adams
that you can rearrange the letters in "overclocker" to spell "clever crook"?
try { do() || do_not(); } catch (JediException err) { yoda(err); }
Thats precisely why Strom Thurman should use 256bit encryption!
0 00000 years... he should be retired by then at least.
2 ^ 256 trials * 1 cycle / trial * 1 second / (10^12 aggregate computer cycles) * 1 year / (3600 * 24 * 365) = 3160000000000000000000000000000000000000000000000
The mention of Pentium-class in a post talking about 750Ghz is almost enough to make me want to throw up.
There has got to be something better than x86. And if consumers are still stuck with x86 when processor speeds hit 750Ghz for the common computer, well, I have lost my interests in computers for life.
I can see it now:
750Ghz PentiumXXXIV processor 1Ghz FSB 2Ghz memory 500Mhz access to storage and a graphics processor that is only capable of pumping out three frames per second in Quake 25.
We have got to leave behind the baggage before we hit multi-Ghz speeds. Please god, don't keep the architecture.
------------
One problem with these high clock rates is that you end up having to pipeline things rather excessively all over the place. I'd imagine at 750GHz that even a single 64-bit ADD would be pipelined over multiple cycles, due to transport delay!
Think about it: Light travels about 1 foot per nanosecond (30cm). At 1GHz speeds, a signal could travel well across a die if it were unimpeded (eg. could travel at the speed of light). In fact, it could theoretically travel most of the way across the motherboard in one clock period. At 750GHz, light travels 0.4mm per clock tick -- about 1/20th the way across a typical CPU die (assuming a die in the range 8mm x 8mm to 10mm x 10mm die -- not too far off what we build today). We're talking 20 pipeline stages just to get from one edge of the die to the other, if we can travel at the full speed of light in a vacuum. And the bad news is that we probably can't -- just look at todays CPUs!
What'll happen is that highly parallelizable problems will speed up, and inherently serial problems will end up staying the same. All of your number crunching for playing video games will rocket along since the calculations can be pipelined and parallelized, but the twisty, turny, five-instructions-and-a-branch control code won't speed up much.
--Joe--
Program Intellivision!
Program Intellivision!
Now we just have to get the price down... Superconductors and 5K cooling systems are both insanely expensive. However, if I can get one of these today, I'd be willing to install the cryogenics facility in my house. That 750Ghz system would be a nice litte boost up from my P133 :)
Joe Consumer -
"750GHZ! WOW! NOW I CAN RUN AOL EVEN FASTER!
AND WITH 56k AOL IS FASTER THAN EVER!"
I just can't see explaining why I need a cryogenics tower for my computer to my wife...
LOAD "SIG",8,1
LOADING...
READY.
RUN
Finally a subject on which I have a decent contribution to make. I wrote a technical report on the technologies behind the current fastest supercomputers and on up-and-coming innovations. This gives a high-level overview of ASCI Red, IBM's Blue Gene, and the HTMT (superconducting technology based) project. Follow this link to the LaTeX2HTML version or download the Postscript version.
"The urge to fly from modern systems, instead of moving through them to even greater, fairer things is, I think, an indi
Yeah, well, the servers crashed because they're cryogenically cooled superconductors, and we ran out of liquid helium this week.
I have to go to the drugstore to buy a few more pounds of liquid helium, I'll be back after lunch.
Uh...thin indium wires are used routinely on any instrument running at very low temperatures to limit heat input. The only way you can get a signal out of an instrument at say 4K and keep the instrument at that temp is to use thin wires.
"When you sit with a nice girl for two hours, it seems like two minutes. When you sit on a hot stove for two minutes, it
The idea isn't new; Congo used a "special form of diamonds" that would be used in exactly this manner as the McGuffin that was the excuse for them to go to Africa. And the book dates back 20 years.
If you're not part of the solution, you're part of the precipitate.
IIRC, isn't Moore's law a relationship between time and transistor count rather than time and speed? While one could find a correlation between transistor count and speed, it's not really as relevant as people think.
:)
I think it could, theoretically be possible (although rather improbable) to reach 750GHz in 5 years and stay right on time with Moore's law, it would just be a matter of cooling.
or maybe I'm an idiot. I dunno. i hope I'm right, because by posting I just lost the ability to moderate this thread
"I hope I don't make a mistake and manage to remain a virgin." - Britney Spears
How 'fast' does our brain work? Can we generalize the synaptic process into a 'Hertz' unit of speed? Can this thing 'think' faster than a human brain? Would it be simpler to tie it right into our lobe than to use something primitive like a keyboard and mouse? Obviously the technology to do so isn't here, yet, but its kinda like sci-fi coming to life...
--
Good quote, too many chars. Seriously, the slashdot 120 char limit sucks!
I've worked for 4 months at the National Research Council in Ottawa, Canada, and there's only one thing you can use to cool down to 5K: liquid helium.
We used liquid helium to cool our experiment. Back in '93 in bulk it costs about $10 CDN per litre. And it evaporates instantly on contact with air. You need to use liquid nitrogen to make sure that the surfaces holding the liquid helium are cold enough so that the helium doesn't just completely evaporate on contact.
I saw another post saying $20k to cool the machine. That might be the cost per month of operation. While super fast chips may be feasable, the most cost effective cold you're going to get is just from liquid nitrogen. I'd probably try to start from there as a benchmark.
(This is the kind of thing I expect to read about some drunk New Zealanders doing in their basement. LHe is just a bit too expensive, I guess...)
If you thought people came up with extreme overclocking methods before, just wait until they try to reproduce this in their garages....
The Free ODMG Project needs volunteers.
Finding God in a Dog
If moore's law holds, and I'm not one to predict whether or not an estimate would fail (who am I to do such a thing?) we should be somewhere around six or seven gigahertz by the time we're all scoffing at this article's headline.
Six or seven gigahertz. We'll be finished simplifying the user interface FAR before then. We'll be fancifying it.
--
Game over, 2000!
I could just see it: Intel starts competing in the microwave business with their new super hot 750ghz chips, which flash fry food in seconds. Hell use a 750ghz bewolfe to heat your office building!
Aside from that I don't see what the point is. Without RAM to match the 750ghz clock speed the chip's value would be seriously reduced. But I digress.
Maskirovka
History is on the move: those who fail to keep up will be left behind. Those who get in the way won't survie at all.
I can get 100,000 frames / second on Q3. Dammit, I can see the difference!!
--
Sometimes it's best to just let stupid people be stupid.