Become a fan of Slashdot on Facebook

 



Forgot your password?
typodupeerror
×

3-D Flexible Computer Chips 85

Roland Piquepaille writes "Engineers at the University of Wisconsin-Madison have isolated a single-crystal film of semiconductor from the substrate on which it is built. Then they transferred this very thin film — 200 nanometers thick — on plastic. Both sides of the film can host active components and several layers can be stacked, opening the way to very powerful 3-D flexible computer chips. Besides computer chips, this technique could be used for solar cells, smart cards, RFID tags or active-matrix flat panel displays."
This discussion has been archived. No new comments can be posted.

3-D Flexible Computer Chips

Comments Filter:
  • The Future is now (Score:1, Insightful)

    by RMB2 ( 936187 )
    ... bringing us that much closer to the roll-up computer screens of Tek War.

    Great show, man. Shatner was amazing
  • by plasmacutter ( 901737 ) on Monday July 24, 2006 @12:32AM (#15767896)
    Heat dissipation is a major issue in cpus.. imagine if you could integrate your cpu within the heatsink's mesh at a monocrystaline level.

    It would be a revolution in cooling efficiency.
    • Heat dissipation is only a significant issue in a very small number of CPUs like x86s and PowerPCs. Most CPUs you'll encounter in your average day (phone, car, mouse, PDA, refridgerator, washing machine, air conditioner, ...) use very little power and you don't heat up that you can notice.
      • Oh? (Score:4, Insightful)

        by SmallFurryCreature ( 593017 ) on Monday July 24, 2006 @02:27AM (#15768085) Journal
        Well my phone does get warm if I use all its features and my iPod sure warms up when it plays video and my PSP and DS both can get a bit warm if played in their case. (Why yes I am a gadget whore)

        Presumably the CPU in larger devices like my washing machine is properly placed and can use the entire body of the machine to cool it, but in general electric circuits of any kind will produce heat under load.

        Saying it just PC cpu's is idiotic. All cpu's will get warm, just because some you use are small enough and cooled well enough that you don't notice it doesn't mean they won't overheat if you remove the cooling.

        • I might be wrong, but the heat from your mobile phone, when using many functions at once, might come from increased energy consumption and therefore higher battery temperature. Same with your iPod, although the conceived warmth of your mobile phone may also result from radiation energy absorption in your skin (only when making a call, tho).
          • Of course you're right in the battery being an added source of heat. Batteries are not that inefficient, so it's in a comparable range to the power usage, which is just about equivalent to the heat put out by the circuit. The actual transmitting power is also not that efficient from an antenna, you get a lot of localized heat in the antenna itself, so if you actually hold the object, I would think that the direct heat transfer through conduction is lower than the radiative absorption.
        • Actually, the Arm9 and Arm7 in the DS run so cool that you could touch them directly and you wouldn't notice the difference between on and off. The heat coming out of your DS is a combination of the battery (~70%) and the screen (~30%).
    • Heat dissapation relies on a large surface area, cpu's need to be small. If you spread a cpu over the area of a heatsink you wouldn't have a heat problem, you would have a slow processor.
    • I wonder why CPUs don't already come perforated for liquid cooling. Instead they rely on thermal conduction to metal and fan/convection. Folding CPUs inside whatever heat transfer medium is used seems a lot of work for the same effect as perforation.
  • So one could possibly make the registers have a "Z" axis and have "real" 3-D address space. Just a thought...

    • "So one could possibly make the registers have a "Z" axis and have "real" 3-D address space. Just a thought..."

      You must have really enjoyed working in a segemented memory environment...
      • Segmented memory is not needed for 3D or even 4D adress space:

        32bit address bus can be divided in 4D 8bit axises a.b.c.d. IPv4 is a 4D address space :)

        Any address bus size can be arbitrary sliced in whatever bit fields, addressing several dimensions/space. Computation penalities applies on non byte/word bondaries though. It depend on chip architecture but spatial chip layout has little to do with that.

        --
        Léa Gris
        • Of course it doesn't, but from my reading of the op it seemed that he/she was proposing that registers have both a physical and conceptual multidimensional component which would imply the use of some kind of offset(s) to address a specific component of said register, no? MOV AX[0][1],... sorta thing
          • Yes, yes... But like writing bits vertically to multi-platter disk drives at the same head location, only with this thought, it could load registers with addresses, or stack, or instructions, whatever. Each vertical location, or part of that bit string, could do something else while the horizontal part is ececuting... I haven't really thought of the possibilities, but both could execute in chain step, even though they would be on different physical processors, or devices of any kind for that matter. There i
  • So does that mean we geeks will finally start wearing $1000.00 clothing just like supermodels? (albeit clothing that runs at dual clock speeds of 6.5Ghz ...)
  • Could be used... (Score:4, Insightful)

    by QuantumG ( 50515 ) <qg@biodome.org> on Monday July 24, 2006 @12:50AM (#15767930) Homepage Journal
    but won't. Call me when the technology is even remotely ready for commericalization.
    • Sounds good... I'll chalk this up after those mega-capacitors we've all been hearing about.

      --nick
    • by rolfwind ( 528248 )
      People probably once had the same attitude toward the transistor when it first came out....
      • No, see, cause the inventors of the transistor were all people who were actually working to make a profit. They weren't research graduates at a university dicking around and writing press releases about how great they are. Now, if IBM or Intel had written this press release I'd be waiting with baited breath for their next great leap forward..

      • Er, no, they didn't. The importance of switching was well established by the time Bardeen/Shockley/Brattain did their work. Computers were already tremendously valuable, and the value of switching from vacuum tubes, which died on the order of a week and were at that time around the side of a lighter, to transistors, which even then could be clustered to a hundred in the size of a penny, was immediately obvious. Their boss at Bell Labs publically stated a week after the discovery that he thought it would
        • Actually, some people did have that attitude - have you ever heard of the vaccuum tube vs. transistor debates in the sound field when transistor radios came out and still go on to this day?

          I didn't say everyone had that attitude, but I should have specified some, and certainly in some niches.

          Guessing doesn't make you smart. Please talk about what you know about only. Please try again soon and come again. Thanks!
          • Actually, some people did have that attitude - have you ever heard of the vaccuum tube vs. transistor debates in the sound field when transistor radios came out and still go on to this day?

            That's a fidelity issue. That has absolutely nothing to do with whether transistors are useful. This is simply an observation that vacuum tubes have a characteristic impact on the domain of sound. Because that impact is reflexive - it's exactly opposite in production than recording - then when music is recorded with tu
  • And thus... (Score:2, Insightful)

    by Null Nihils ( 965047 )
    Moore's law will be able to keep marching on.

    Hopefully these tricks, and others, will be commercially available by the time it becomes impractical to cram any more cores on a single chip.
    • The Terminator is yet another step closer to reality.

      In seriousness, this is very exciting news. I wonder how difficult it will be to automate the extraction process.

      We need to combine this with oleds and power paper to make a complete device that is the size, shape, and thickness of a
      piece of paper. Add flash memory and you could literally have a notebook full of computers. If the tech is fast enough,
      it could even be run low voltage and have no need for any external heat management under normal operatin
  • This is new .. how? (Score:5, Informative)

    by Bender_ ( 179208 ) on Monday July 24, 2006 @01:15AM (#15767977) Journal
    This is not new - this is known as "susbtrate transfer process" and has been practiced for year. One company doing very advanced work in this is Philips:

    First two paper hits I found in google:

    http://retina.et.tudelft.nl/data/artwork/publicati on/hf/ectm013.pdf [tudelft.nl]
    http://retina.et.tudelft.nl/data/artwork/publicati on/hf/111568631.pdf [tudelft.nl]

    Many companies are also working on substrate transfer processes to build silicon wafers with selective crystal orientation. Among them IBM and Soitec.
    • by Anonymous Coward
      Did you read the articles or the pdf's you sent? These are completely different processes. Not only are the products of the transfer very flexible in the new process but they can also be stacked which is certainly not achievable in the second and maybe not in the first, although it doesn't specifically say. Are you just trolling?
    • this is a link to their JAP abstract [aip.org]

      it seems that the novelty is that they were able to transfer strained silicon thin films to a polymer substrate.

  • Does anyone have any ideas, besides clothing, as to how this could be useful? As I see it most devices have a hard screen, which makes them inherently inflexible.
    • Some sort of intelligent condom?
    • Comment removed based on user account deletion
    • How about animated tattoos [howstuffworks.com]? I know it's a coupla rednecks who put this together (and filed a patent), but I didn't quite think they put together a flexible LCD by themselves...
      • Not only didn't they do it without help, they didn't even do it at all. Did you read the little box at the bottom?
        The patent for the PSVI is real, but that's as far as it goes. Now that April 1st has come and gone, we'll admit that this article is absolutely untrue...for now.
    • The reason they have hard screens is that the edge feeding circuits are inflexible. This is a reasonable, real-world flexible circuit. LCD and OLED screens aren't by their nature rigid; they're just plastic bubbles full of goo. This is (other than scratch resistance and the power supply, both of which are solvable but ugly) the last major barrier to rollable screens.
  • Fire (Score:4, Funny)

    by luketheduke ( 945392 ) on Monday July 24, 2006 @02:05AM (#15768051)
    In other news Geeks overclocking clothes causes fire and personal injury.
  • Imagine (Score:5, Insightful)

    by cryptoluddite ( 658517 ) on Monday July 24, 2006 @02:13AM (#15768064)
    They say this can be used for solar cells too. Imagine what 300 billion dollars investment could have done to make this a reality. We're like the 3rd generation of rich kid, the one that pisses away the fortune on gambling and yahts instead of doing something productive with it.
    • So you had rather that the family cycle didn't exist, and the entire world was by now owned by one man? At least the third generation rich kid provides a living for Indian tribes and interesting employment for other people.

      Being serious, your argument is flawed. This might not in fact be a good way to make better solar cells. You can invest as much as you like in technology, but if you try to push too far in one direction too fast you will fail to get synergies. Putting a man on the moon has actually achiev

      • > So you had rather that the family cycle didn't exist, and the entire world was by now owned by one man?

        That's soooo flawed. What about the whole world sharing the wealth where it's needed and not in the 20-80 80-20 proportions?

        > Before you can spend $300 billion on R&D you have to get a big enough educated population,

        Spend half of that $300 billion on the education. Actually, don't separate education from research, simply make one morph into the other with experience.
        • "What about the whole world sharing the wealth"

          Not gonna happen... wealth is not how much you have, it's how much /more/ you have. A closer gap between rich and poor will mean that money will have less value overall.

          • Actually, the value of any commodity is based upon the quantity of that commodity you have. The more of a commodity you have, the less value it has to you. The less of it you have, the more each unit is worth. You've got it backwards.

            Economics 101...
            • When you're 100% self sufficient and have everything you could ever want/need, maybe, but be realistic, is that gonna happen often? No, so you need to make exchanges, and the *exchange* value of anything is based not on how much you have, but how much more you have than someone else (with something that you want), hence their desire to exchange with you, actually. For example, if everyone starts getting paid more money, the money starts to lose it's value; this is called "inflation". I don't have it backwar
              • You didn't mention self-sufficiency, nor did you mention inflation, which, by the way is NOT everybody getting paid more. See this: http://en.wikipedia.org/wiki/Inflation [wikipedia.org], where wikipedia gives a more proper definition, that's not so self serving as yours. Strictly speaking, inflation is too much money chasing too few goods. I guess your mentioning money loosing its value is the closest you came to being right, but that's not because everybody gets paid more. Everybody getting paid more is only a sympt
                • Don't accuse me of obfuscating just because you can't understand what I'm saying, and you're blind if you don't think that increased wealth doesn't drives up prices. Wages *must* increase inline with inflation, otherwise a recession will occure as people can afford to buy less than they could previously. But just because increased wages is in effect of inflation, does not mean that it's not also a cause. It's called a "cycle"; two things that are both cause and effect of each other.

                  My original statement, wh
          • Meaning it's good if many people are poor because then money have better value then?
            Something's wrong about the priorities here.
            • hey I never said it was good, but it is human nature. When it comes to the crunch, people will prioritise their own families needs above others (and other times just out of greed) and so will try to maximise what they get out of each exchange made.

              For example, when people shop around for "the cheapest" (eg, go to a shop where banana's are cheaper), the result being that people who make the produce get paid less.

              When money's tight, I regret I must admit I do exactly that, as do most, and if you ever have don
              • "when money is tight".
                The OP was about people for whom money is definitely NOT tight.
                Agreed, people already understood leadership based on inheritance (monarchy) is not the way to go and replaced it with sharing equal power of decision who's to rule, aka democracy. Now if the same happened to economy...
                • "replaced it with sharing equal power of decision who's to rule, aka democracy. Now if the same happened to economy"

                  Corruption would occure just as it does in a "democracy", and where corruption doesn't occure, things will often end up being driven into the ground because decision is given equally to people who are less able to make good judgement. Various levels of this can occure with economy, such as socialism (the rebalance of some wealth) or communism (wealth spread evenly), but human nature stops them
            • Nobody said that. My statement was a bald faced explanation of an economic axiom about the marginal value of money, or for that matter any commodity, NOT any kind of moral judgement as to the goodness or not of that axiom.

              What the axiom means is that, as you accumulate more of anything, each additional unit of that commodity becomes less valuable. If you have ten bucks to your name, five more dollars is an increase of 1/3 of your net worth. If you have $100,000.00 its less than pocket change. The same
    • Comment removed based on user account deletion
    • We're like the 3rd generation of rich kid, the one that pisses away the fortune on gambling and yahts instead of doing something productive with it.


      Gambling is zero-sum; nothing is created or destroyed. It's merely a mechanism for transferring money from people who are bad at math to the mafia. It therefore cannot constitute 'waste' at this point (what the recipients do with it is unclear).
  • by Anonymous Coward
    He's Back! Back from vacation, fresh faced and fancy free! Ladies and gentlemen, I give you MR. ROLAND PIQUEPAILLE! Give him a big hand!

    (by the way, I spell-checked this in Word. PIQUEPAILLE was not flagged as a bad word... ???)
  • by Anonymous Coward


    victoria's secret is now a hardware store...

    "no, hon, that wasn't your manliness: my bra's cpu is just overheating again"

    ... and myspace just got that much more obnoxious.
  • One step closer to my Iron Man armor. Call me crazy, but the tech is not so far-fetched.
  • "Beer bottles that use solar power to keep their precious contents cool in the height of summer could be a welcome fringe benefit of thin-film technology currently under development"

    rest on http://www.vnunet.com/2160151 [vnunet.com]

    I'm pretty sure this wouldn't work in direct sunlight, as I doubt the heat pumps and solar cells are efficient enough. In diffuse light, which an ideal application of the less efficient thin film cells, it might work quite well. Haven't done the maths though.
  • Now they've got the semiconductoroff the substrate and flexible, as strained monocrystals, they should be able to align corresponding sites on the multiple layers of circuits to actually go 3D, for more efficient routing. Maybe even rolling up sheets into scrolls. Since they're flexible, maybe rolling scrolls around a power core, making tiny smart wires.
  • Carbon nanotube networks do this faster and cheaper, have already been used in solar cells, and are being commercialized right now.

A morsel of genuine history is a thing so rare as to be always valuable. -- Thomas Jefferson

Working...