Become a fan of Slashdot on Facebook

 



Forgot your password?
typodupeerror
×
Hardware Science

PC's Waste Heat Could Add To Processing Power 134

Urchin writes to tell us that physicists working in a new field called "phononics" claim that waste heat from a processor could actually be used to add to its power. "Crunching data coded using photons — photonic computing — is one example, and in 2007 researchers built the first workable optical transistor. But now the idea of computing using heat flow is gaining popularity among applied physicists. Heat travels through solid materials by means of phonons — ripples of vibration passing through a series of atoms. Those ripples can be used to send and store data in digital form: one temperature is read as 0 or 'off' while a second, higher temperature is interpreted as 1 or 'on.' Provided that the thermal memory is well insulated, it can keep its temperature — and data — intact for a long time."
This discussion has been archived. No new comments can be posted.

PC's Waste Heat Could Add To Processing Power

Comments Filter:
  • Nice, but... (Score:2, Interesting)

    by samriel ( 1456543 )
    I might not entirely understand this, it sounds like this is a whole lot of work for not much result. What happens if you get temperatures that are precisely inbetween 0 and 1 values? What effect does the processor's fan and/or heatsink have on said values? Why bother?
    • Most of the time it's a threshold. So if it's, say, 50*C or above, it's a 1. Anything below that is a 0. They do a very similar thing with off/on states for memory and such. It's a threshold, not a completely discrete value.
      • Re: (Score:3, Interesting)

        Didn't read TFA but it could be a hysteresis [wikipedia.org] kind of thing which is thought of as a boundary region rather than a single threshold. Or, like a Schmitt Trigger in electronics, an upper and a lower threshold to eliminate jittery triggering.

        If you have a noisy, fluctuating signal then it makes sense to output a 1 when the upper threshold is crossed, and to output a 0 when the lower threshold is crossed.
        • by hdon ( 1104251 )

          Didn't read TFA but it could be a hysteresis [wikipedia.org] kind of thing which is thought of as a boundary region rather than a single threshold.

          If you have a noisy, fluctuating signal then it makes sense to output a 1 when the upper threshold is crossed, and to output a 0 when the lower threshold is crossed.

          You mean 0 when the threshold is crossed one way, and 1 when crossed the opposite way?

          • Comment removed based on user account deletion
          • No, you have two thresholds. The region in between them is indeterminate, while the region above the upper threshold is 1 and the one below the lower threshold is 0. The distance between the two could be determined based on the signal-to-noise ratio.

            At least, I'm pretty sure that's what the GP was getting at.

    • What happens if you get temperatures that are precisely inbetween 0 and 1 values?

      I'd assume the thing that happens on an electrical signal path when you get a voltage that's precisely inbetween 0 and 1...

    • Then it fails the cyclic redundancy check and has be be ...reheated I guess.

  • CPU Turbo (Score:3, Interesting)

    by pak9rabid ( 1011935 ) on Friday January 30, 2009 @03:10PM (#26670677)
    Interesting...kind of like a turbocharger for a CPU.
    • Re: (Score:3, Funny)

      by Anonymous Coward

      Sounds like there will be a new market for all of those 386 boxes with the turbo button.

    • Yeah, but now we'll start seeing VTEC and Type-R badges too...

      • FYI, VTEC isn't really the same as a turbo charger. While it does give you more power when you need it, the real benefit is making the ride smoother at higher speeds by lowering the RPMs needed to maintain the speed. See here [youtube.com] for a short video on it.
        • by tom17 ( 659054 )

          Lowering the RPM to maintain the speed? You know, changing up a gear has that effect, right?

          I don't see how adjusting the lift profile of the valves can affect RPM vs Speed...

        • FYI, VTEC isn't really the same as a turbo charger.

          The people to whom he's alluding [imageshack.us] don't care.

        • Re: (Score:3, Informative)

          by Aranykai ( 1053846 )

          VTEC stands for Variable Valve Timing and Electronic Lift Control. It allows the engine to adjust the valve timing in real time to provide additional power at higher RPMs. It essentially allows low displacement engines to expand their powerband at very high RPMs.

          It does absolutely nothing for ride quality and can be employed at any road speed, provided the transmission allows the engine to operate within the RPM required for VTEC to operate.

          • Re:VTEC (Score:3, Insightful)

            To be fair, VTEC does allow for a milder cam profile at lower rpms, while allowing for a much more aggressive cam profile for additional horsepower at higher rpms. Maybe that's what the OP was getting at.
      • by DittoBox ( 978894 ) on Friday January 30, 2009 @04:22PM (#26671511) Homepage

        With such memorable phrases as:

        "See these stickers? They make my computer go faster!"

        "With this giant wing on the back of my front-wheel drive computer I get the down force I need to go fast"

  • by jellomizer ( 103300 ) on Friday January 30, 2009 @03:10PM (#26670679)

    Lisa, In this house we obey the laws of thermal dynamics!

    That said. It may save some power converting loss head back again making it more efficient.
    But they way that most people use computers I don't know if there is a benefit. We rarely run at full CPU Heat kicking.

    • We rarely run at full CPU Heat kicking.

      That's the 2nd time I've heard that this month.

  • by zappepcs ( 820751 ) on Friday January 30, 2009 @03:11PM (#26670685) Journal

    I wonder how long before this is used for something bad? Does this possibly mean that the sun, inhabited by an alien life form, has turned off the one's and zero's in an effort to relay the message GTFO!

    FTFA:

    Casati says practical physicists must rise to the challenge set by the theorists. Yet even if they can, phononic computing is unlikely to threaten electronics because phonons travel a lot slower than electrons. Li imagines that the two technologies will work together, in hybrid devices that perform some computation using waste heat.

    I bet there are better ways to use this than PC computing

    • by Rayeth ( 1335201 )

      There are easily better applications, but you'd be surprised how much uses "PC Computing" these days.

      How about enabling automatic de-icing systems on planes without needing pilot intervention. Or using the engine exhaust heat to tune your car's performance?

      All of these use computers and are totally valid applications of TFA. So I hope they develop these devices if only for use in other computing fields.

    • by morcego ( 260031 )

      I bet there are better ways to use this than PC computing

      Not only that. Since this phononics computing has absolutely no advantage over conventional (electronic) computing, the amount of time and money wasted is ... well ... a waste.

      There are several viable options here, and I bet most, if not all, are already under study and development:
      - Improving the efficiency (less power waste due to Joule effect )
      - Converting this back into electricity (yes, it is not very efficient, but it is better than wasting)
      - Us

  • Hooked on (Score:5, Funny)

    by Anonymous Coward on Friday January 30, 2009 @03:11PM (#26670689)
    I'm hooked on phononics. And quit making fun of my stuttutter!
  • Hmm (Score:5, Insightful)

    by ZwJGR ( 1014973 ) on Friday January 30, 2009 @03:12PM (#26670707)

    "Provided that the thermal memory is well insulated", that basically means putting it on a different piece of silicon/on something else entirely, which kind of defeats the object as I see it.

    While I haven't looked at this in great detail, it strikes me that achieving anything near useful density is going to very difficult due to entropy, and the simple fact that putting very small volumes at slightly different temperatures right next to each others quickly leads to a relatively uniform temperature distribution.

    This sounds somewhat improbable/unfeasible to me...

    • Re:Hmm (Score:5, Funny)

      by evanbd ( 210358 ) on Friday January 30, 2009 @03:26PM (#26670865)
      Quick, someone tell the physicits! I'm sure they forgot all about this. Good thing we have /. to check their work.
      • Re:Hmm (Score:4, Insightful)

        by morcego ( 260031 ) on Friday January 30, 2009 @04:43PM (#26671753)

        Just because the math works, doesn't mean it is viable to implement.

        As TFA say, the theorists just came up with the idea and some of the math, and pretty much left to the practical physicists to find a way to implement it.

        Next, how to use your farts, produced while using a computer (wasted right now), to increase its power.

        • Next, how to use your farts, produced while using a computer (wasted right now), to increase its power.

          Ummm....my system is overclocked right now.

        • Next, how to use your farts, produced while using a computer (wasted right now), to increase its power.

          I'm channelling mine to Washington. Each fart produces more brainpower than Capitol Hill produces in a week.

        • "practical physicists" - didn't they use to call them engineers?

      • Re: (Score:2, Insightful)

        by tyrione ( 134248 )
        No offense, but physicists always leave it to engineers to come up with a viable solution.
      • Re:Hmm (Score:4, Interesting)

        by Vellmont ( 569020 ) on Friday January 30, 2009 @07:55PM (#26673715) Homepage


        Quick, someone tell the physicits! I'm sure they forgot all about this.

        More like someone tell the journalists (the people who actually WROTE this article). It happens all the time that a scientist says something offhand like "and you could use this for processing power", and a journalist misinterprets this to mean that it's both feasible, and commercially viable.

    • heat to a certain point, the joint points melt, connecting two neighboring leads, and congratulations, you have a short... hooray for phononics transistor!

    • by hdon ( 1104251 )

      "Provided that the thermal memory is well insulated", that basically means putting it on a different piece of silicon/on something else entirely, which kind of defeats the object as I see it.

      To interface a phononic computational device with a traditional electronic device without leaking heat, you could use something like an optoisolator, though obviously it might not be preferable to use an electric photo element.

      While I haven't looked at this in great detail, it strikes me that achieving anything near useful density is going to very difficult due to entropy, and the simple fact that putting very small volumes at slightly different temperatures right next to each others quickly leads to a relatively uniform temperature distribution.

      From the article:

      It exploits the fact that some materials can only exchange heat when they are at similar temperatures.

      This sounds somewhat improbable/unfeasible to me...

      You're probably right, it probably is infeasible. But then again, a lot of good things do start out that way.

    • I'd assume it would have to operate at pretty high frequencies constantly being refreshed or not at all.
  • by loose electron ( 699583 ) on Friday January 30, 2009 @03:13PM (#26670715) Homepage

    This, IMHO is an academic concept at best. State definition by thermal state has been done in research before but it is slow, and trying to collect the waste energy in the form of heat and re-use it as the byproduct in another state machine sounds a bit questionable.

    Mechanical computers are viable as well, but not too terribly practical.

    • by Chabo ( 880571 )

      Mechanical computers are viable as well, but not too terribly practical.

      As proved in 1991. [wikipedia.org]

    • Mechanical computers are viable as well, but not too terribly practical.

      That's largely a question of scale, isn't it?

      • In theory - if you can use a medium to create a NAND gate, then anything and everything needed to create a computer can be defined from that. Again "in theory" is the important catch phrase. You can do that using relays, transistors, vacuum tubes, thermal storage, mechanical logic, and water valves. All have been demonstrated. The practical realization of a big scale machine is another issue.

  • Yes, temperature gradients are a form of information, but get real guys: they are the bottom of food chain of useful computing tech. A real computer dumps heat as a side-effect of doing useful work - its more efficient to try to recover the heat as energy than to directly use it to compute. Sure, if you have a steam locomotive, it makes sense to add secondary and even tertiary energy-extractors to increase locomotion, but that is a special case: the locomotive is isolated and has big energy needs - this

    • Yes, temperature gradients are a form of information, but get real guys: they are the bottom of food chain of useful computing tech. A real computer dumps heat as a side-effect of doing useful work - its more efficient to try to recover the heat as energy than to directly use it to compute. Sure, if you have a steam locomotive, it makes sense to add secondary and even tertiary energy-extractors to increase locomotion, but that is a special case: the locomotive is isolated and has big energy needs - this is not the case for computers.

      I like your idea about trying to recover the heat. The only serious proposals I've heard involved using it to heat the building in which the machines are stored during cold weather. I'm a bit surprised that I haven't heard of anyone using something like a Sterling engine (maybe these could be miniaturized?) to try and do useful work with the waste heat, in much the same way that electric cars can use regenerative braking to minimize wasted energy. Or perhaps there could be a device like a solar panel tha

      • would be interesting (Score:2, Informative)

        by zogger ( 617870 )

        Seems like a solid state thermocouple might be easier to use. I'd like to see some sort of heat pipe from the case to one, then use that output to power the screen (maybe not the main one but a smaller backup little screen??). I have no idea of the state of the art there though or what sort of useful electricity you might get from one. I have seen a kerosene lantern from Russia that uses a thermocouple to scavenge waste heat from the kero burning to provide power for a table radio.

      • Re: (Score:3, Informative)

        by Gorobei ( 127755 )

        At one of our compute farms, we actually pipe the waste heat into a local town as a low-cost house-heating solution (think steam-pipes, but lower quality energy.) It works there because even 100 degree hot air is nice to have when the outside temp is 0f.

      • Your ideas are intriguing to me and I wish to subscribe to your newsletter.

      • The only serious proposals I've heard involved using it to heat the building in which the machines are stored during cold weather.

        Doesn't this naturally happen just by running the machines in the building?

        • The only serious proposals I've heard involved using it to heat the building in which the machines are stored during cold weather.

          Doesn't this naturally happen just by running the machines in the building?

          To some degree, yes. I think those proposals had more to do with collecting the heat and integrating it into the central heating system of the building. In other words, it may be more of a problem of efficient distribution (maybe only one room or a few rooms have lots of computers and you want to heat all of the building, for example). I'm using a lot of words like "I think" and "it may" because I'm honestly rather ignorant about these systems, but I believe that's the gist of it.

  • I have this picture in my head of a water-cooled PC that gets so hot that the water turns to steam, which runs a turbine, which helps power the PC.

    Hugely inefficient, but sooooo cool to have a Steam-Powered PC.
  • Who ordered that? And where do they fit in the standard model?
    • Re: (Score:3, Interesting)

      by MarkusQ ( 450076 )

      Who ordered that? And where do they fit in the standard model?

      Phonons are just a very weird state of photons, for suitably high values of "very weird"--they propagate by the dipole interactions of the substrate, and are thus at the bottom an EM phenomena. But that isn't a very useful way to look at them (about as useful as saying that sound in air involves oscillatory motion of masses and thus could be considered a source of gravity waves).

      Rather than asking "how do they fit into the standard model?" it

  • Tooooooo slooooow. (Score:5, Insightful)

    by goodmanj ( 234846 ) on Friday January 30, 2009 @03:31PM (#26670917)

    Phonons travel at the speed of sound in their medium, which is 100,000 times slower than the speed of electrical signals or light. If you've got a phononic circuit running at a Ghz clock rate, signals can only travel a few microns. This size limit severely restricts the number of individual components you can have in your circuit.

    Go light, or go home.

    • by durrr ( 1316311 )
      For those who missed it; the overall utility of something is seldomly measured only in the FPS it achives in crysis.

      The utility of phononic circuits stretches into areas where most electric circuits would not thrive. being able to process information with the only requirement being a temperture gradient would be quite handy for many occasions, considering you usually have the body-surroundings temperature gradient availible.

      Even ignoring any direct applications we're likely to see within a few decade
      • by fyoder ( 857358 )

        The utility of phononic circuits stretches into areas where most electric circuits would not thrive.

        For example, Hell, where things have all eternity to happen.

        "Satan, my laptop is taking forever to boot!"

        "Yes, and by the time it's finished, Windows will insist on telling you that all the thousand icons on your desktop are stale and should be deleted! And there will be a ton of updates that have to downloaded and installed! Bwah ha ha ha ha!"

        Much like those rare occasions when I fire up Windows in a virtual machine. You know they use Windows in Hell.

      • by wdef ( 1050680 )
        The speed of transmission between two synapsed neurons in the human brain is about 1mS. But 1mS is a really long time compared to eg nanosecond instruction seek times in RAM. Yet look what the human brain can get up to, even owned by the average Slashdotter. As Durr says, it's not solely about speed.
    • Plus I'd think it would be very sensitive to thermal noise or excessive heat.

  • Just use the heat to create more electricity to increase the efficiency. Thermal Acoustic Piezo Energy Conversion [wikipedia.org]

  • Talk about slow... matter has this undesireable property of "thermal inertia". Your basic thermal wire is going to have a hard time making very many transitions per second. We're talking centihertz.

    They'll have to speed up Vista by a factor of several gazillion to run well on this CPU.

    • by Hatta ( 162192 )

      Thermal inertia? Inertia is the tendency for the position of an object to keep changing at the same rate, unless an outside force is applied. Would thermal inertia then be the tendency for the temperature of an object to keep changing at the same rate, unless an outside force is applied?

      • Replace /position/ with /velocity/ and the principle applies.

        • by Hatta ( 162192 )

          So:

          Inertia is the tendency for the velocity of an object to keep changing at the same rate, unless an outside force is applied.

          I don't think that's correct.

          • no but it does make for an interesting daydream. Just think how fast I could get my car going if it followed the new and improved version of the first law of motion!
  • This is how the outer layers of a Matrioshka Brain would work, perhaps?

  • So conceivably, you could have patterns of heat moving through a large mountain, or a lake, or a large gas field in space, constituting a conscious intelligence.... hrm.
    • by osu-neko ( 2604 )
      Well, if we keep finding new ways to encode information and to engage in computation, it's conceivable that someday, everything in the universe could be used for computing. Essentially, every last bit of matter and energy will be part of a Beowulf cluster. ;)
      • Well, if we keep finding new ways to encode information and to engage in computation, it's conceivable that someday, everything in the universe could be used for computing.

        And we can use it to answer The Last Question [wikipedia.org].

  • AMD stock jumped suddenly today.

  • Thermos porn...I can't wait!
  • What if the universe is just binary code.
  • The net gain in computing power would be negative, simply because everyone knows that heat causes computrons to decay into bogons, resulting in an overall loss of processing power.

  • I didn't realize PC's produced dung, let alone that this dung produced much heat! I've seen my PC goto shit, though.
  • Caller: Hi, my computer keeps getting these blue screens.
    Tech: Have you recently installed new software or hardware?
    Caller: No, it just randomly started getting blue screens.
    Tech: Have you had any power outages lately?
    Caller: No, everything has been fine up until now.
    Tech: Have you tried turning it off and on again?
    Caller: Yes, I even turned off my printer and my heat lamp and speakers and all the other stuff plugged in here.
    Tech: Sir, I'd recommend you turn off all the lights in your room while you
  • by beef curtains ( 792692 ) * on Friday January 30, 2009 @04:29PM (#26671593)

    If I understand this properly (and it's not 100% guaranteed that I do), this sounds like an excessively complicated solution that would yield relatively little benefit. The "sandwich" idea from TFA sounds especially counterproductive, if external power is required to keep the hot side hit & the cold side cold.

    Instead of trying to harness waste heat to eke out a fraction of a percent of extra processing power, here's an idea: how about sucking that waste heat into a small insulated pipe with a low-voltage van, and running that pipe down to my feet? It's very cold near the floor of my apartment, and some warm air aimed at my tootsies would be greatly appreciated while I use my computer.

    Maybe this pipe could have a little door I could close in the summer, when the additional warmth would be less welcome.

    • Re: (Score:1, Insightful)

      by Anonymous Coward

      Could just put your PC under your desk...

  • I hope they are not too worried about interfacing to either quantum http://en.wikipedia.org/wiki/Quantum_computing [wikipedia.org], photonic http://en.wikipedia.org/wiki/Photonic_computer [wikipedia.org], or spintronic http://en.wikipedia.org/wiki/Spintronics [wikipedia.org] circuits. The thermal 1's and 0's would cause immediate decoherence in the other computational circuits, and they have much more potential bandwidth than any thermal type circuit will ever have. Quantum computers will beat the pants off of any thermal flow logic, providing that the p
  • In the winter time (and where I am, it's damned cold out during the winter...snow banks are over my head right now), I have to shut off the heat vent in my office.

    The heat thrown my all the machines in here, one main dev box, 1 laptop and 3 small appliance servers, plus all the peripherals keep the office nicely heated all winter.

    So...what waste heat...I use mine, thanks!

  • Why does a CPU emit heat when X instructions are made? Is there a reason, or perhaps a physical law that requires X quanta of heat per Information instruction?

    I read elsewhere that the waste heat is the result of doing irreversible math on the CPU, and the thrown away information convert to heat. And I saw people working on CPUs that are reversible, which could recover the energy back out of them (minus negligible running costs).

    Is that just pie in the sky academic research, or a viable path for CPUs to eve

    • Re: (Score:3, Informative)

      by meringuoid ( 568297 )
      Why does a CPU emit heat when X instructions are made? Is there a reason, or perhaps a physical law that requires X quanta of heat per Information instruction?

      A computation is a process in which we take a memory area that may be in any randomly-chosen state, and reconfigure it to be in one specific state, corresponding to the return value of our computation.

      This is a local reduction in entropy - reconfiguring that memory area into a single state out of the many possible. That means work has to be done,

    • Re: (Score:3, Informative)

      by artor3 ( 1344997 )

      At the lowest possible level, we're talking about electrons moving around. Every time you do an instruction, a bunch of electrons have to move from one place to another. On the way, they inevitably bump into things. Whenever that happens, you lose a bit of energy as heat. That's what the oh-so-common equation, P=(I^2)R means. The I is the current (moving electrons), the R is the resistance (things electrons bump into), and the P is the power (energy per second that you lose as heat).

      As for what you rea

  • Seasonal (Score:4, Funny)

    by rossdee ( 243626 ) on Friday January 30, 2009 @07:05PM (#26673375)

    Its winter. There is no such thing as 'waste heat'. Every watt emitted by a computer is a watt that doesn't have to be emitted by the heater.

  • ...phononics proposes transmitting data as heat (thermal 'waves', whatever). So the heat generated by losses in various circuits on the chip would be noise to the data phonons we want.

    This is analogous to signal to noise problems in electronic systems. And the solution in the electronic world, reducing the noise, means that we've got to be even more concerned about reducing the heat due to losses in a phononic chip.

Do you suffer painful hallucination? -- Don Juan, cited by Carlos Casteneda

Working...