Slashdot is powered by your submissions, so send in your scoop

 


Forgot your password?
Close
typodupeerror
×
Hardware Science

NASA Wires Chips With Nanotubes 196

Posted by timothy from the little-tiny-ones dept.
carstene writes "SpaceDaily reports that NASA has come up with a way to wire microchips with nanotubes instead of copper interconnects. Aparently this could keep Moore's law a reality well into the next decade."
This discussion has been archived. No new comments can be posted.

NASA Wires Chips With Nanotubes More

NASA Wires Chips With Nanotubes

Comments Filter:

  • royalties (Score:5, Insightful)

    by TerraFrost ( 611855 ) on Wednesday April 16, 2003 @01:38AM (#5742192)
    with the money NASA can get off the patents for these, the space program may indeed have a future! :)

    • Re:royalties (Score:5, Interesting)

      by bcwalrus ( 514670 ) on Wednesday April 16, 2003 @01:42AM (#5742202)
      Can they patent something created using your tax dollars?

      • Re:royalties (Score:1, Informative)

        by Anonymous Coward
        technically speaking they can patent it and license the technology or make it public domain.
      • Yes, they can, besides if NASA gets income from royalties that means (theoretically) less tax dollars will be needed.

      • Re:royalties (Score:3, Interesting)

        by Mac Degger ( 576336 )
        I think technically they can, but I don't think they do.
        I've seen plans for tiny scanning lasers (for docking alignment) on one of nasa's many websites, and loads of other stuff to boot. There's also a host of other reserach papers available online.
      • I hope they do, so companies will have to pay to use this technology created by my tax dollars, rather than gaining the benefit without giving anything back. The space program deserves a chance to make a little money and achieve some sort of semblance of self-sufficiency.
        • Well, theoretically, and probably in practice, the companies will have lower costs, the savings from which they will be able to pass onto you, the consumer, if they don't have to pay licensing fees. Since Intel and AMD are in pretty steep competition, I'd imagine the price will reflect it (if and when these actually hit the consumer market).

          • Re:royalties (Score:2, Insightful)

            by theRiallatar ( 584902 )
            Riiiight. Keep on believing that. We're going to pay for "R&D" costs from companies making these products whether the government developed it or not. Prices are going to be outrageous when this stuff starts to hit the market regardless of how it was developed.
        • If I'm not mistaken, companies in America pay a far higer percentage of the taxes than Joe Consumer.
          • If I'm not mistaken, companies in America pay a far higer percentage of the taxes than Joe Consumer.

            You're not mistaken; you're misled.

            Comapnies are taxed only on their profits--that's like making the food that I buy and the rent I pay tax deductions.

            Corporations also pay "it's not profit" gains. MS, for example, makes an ungodly ammount of money--but since they never declare it as a profit, they don't pay taxes on it.

          • Re:royalties (Score:3, Informative)

            by ShavenYak ( 252902 )
            You are mistaken. According to the IRS [irs.gov] (who should know), in 2001 individual income tax accounted for over $1 trillion in revenue. Corporate income taxes accounted for less than $200 billion.
    • Did they ever patent velcro out of curiousity? If they did that must have got them something, if they didn't then "Doh!"
      • NASA didn't invent Velcro. For an interesting look into the origins of Velcro look here [velcro.com]

        The name Velcro is a contraction of "velour" and "crochet". The loop side is the "vel", and the hook side is the "cro".

    • Apparently, NASA botched the metric conversion and the nano tubes are actually 6 inches across.

  • NASA vs. Intel (Score:5, Interesting)

    by traskjd ( 580657 ) on Wednesday April 16, 2003 @01:40AM (#5742196) Homepage
    Just wondering - but how much would NASA have spent to find this out? I mean It's common to see companies like IBM come up with stuff that is cool like this (like the copper idea a few years back). It seems to me that Intel doesn't actually come up with too many new ideas? (I mean sure there chips become faster but not amazing new things).

    I could be wrong. Has Intel done anything this cool? Surely they would spend more money on R&D for processors (I would assume NASA spends more on Space?)

    any info about this would be much appreciated.

    • Re:NASA vs. Intel (Score:2, Insightful)

      by Tailhook ( 98486 )
      Has Intel done anything this cool?

      How about creating the first microprocessor [intel4004.com]? That cool enough for you?
    • nasa's been in the nano game for a long time (since at least 1996 [nasa.gov])

      http://www.nano.gov/2003budget.html [nano.gov] will give you an insight into the actual dollars.

    • Re:NASA vs. Intel (Score:5, Interesting)

      by Slowping ( 63788 ) on Wednesday April 16, 2003 @02:12AM (#5742288) Homepage Journal
      Intel has always held a very conservative line regarding research into far-out new technologies. The vast majority of Intel's research money goes into fab/mount/production technologies.

      For stuff like nano-tubes and quantum computing, Intel usually helps fund academia to let them take the high-risk endeavors. And then take the benefits as they are produced.

      I'm not sure what Intel's current plan is now, but it seems that they're putting more R&D dollars into the mobile/ubiquitous computing market, to try and branch out their chip options, instead of being forever racing against Moore's Law.

      Intel-research.net [intel-research.net], for some info on Intel and partnerships with academia on this type of research.

    • How about Hyperthreading? That's kinda cool, at least it uses "hyper" in the name like hyperdrive, hyperspace, hyper girlfriend and other nifty things...

      • Re:NASA vs. Intel (Score:3, Interesting)

        by arivanov ( 12034 )
        It was bought with the Alpha team. Dec, then Compaq has had it in development for circa 10 years before don Cappella decided that they cannot make processors (can someone finally deliver him some clue through the relevant orifice). IBM got it through an older partnership with Compaq/Alpha that predated the sale to Intel.

        Anyway, if not this idiotic decision to sell your crown juvels the game in the server town would have been quite different now. Basically the PPC and Alpha would have been multithreaded whi

        • BullSh*t (Score:3, Interesting)

          by djohnsto ( 133220 )
          Hyperthreading was built into the very first Pentium 4 (Willamette) processor, who's design started around 1994 (maybe 1995, I don't remember) under the codename P68. Regardless of when the design started, the first chip was released in what, 2000? That's well before Intel hired most of the Alpha team and got IP rights to Alpha technology.

          Now, I'm not saying the Intel invented SMT (hyperthreading), but they didn't really just take it from Compaq either.

          Most of the Intel inventions are either not disclos
        • Nonexistant products always perform better then real products.

          You can't prove that PPC or Alpha would currently be performing better then Intel for the same price (a critical point; who give a fart that Alphas run more quietly and twice as quickly if they cost 20 times as much (a rhetorical number, BTW, but I believe they were significantly more expensive)?) unless you build one.

          Even if Alpha or PPC had been put on life support it is likely that Intel would have continued to dominate because Intel had a b

  • quantum entaglement (Score:4, Insightful)

    by astafas ( 232064 ) on Wednesday April 16, 2003 @01:43AM (#5742204)
    What interested me more, was that at the bottom of the article, it mentions that we have quantum entanglement of 3 electrons working. I don't know what will be more useful to continue Moore's Law, the nanotubes or the quantum computers. The nanotubes seem to be an evolutionary upgrade where the quantum computers seem to be more revolutionary.

    • Re:quantum entaglement (Score:5, Insightful)

      by addaon ( 41825 ) <addaon+slashdot @ g m a il.com> on Wednesday April 16, 2003 @02:10AM (#5742282)
      Well, in terms of keeping moore's law, clearly the evolutionary technology is preferred. One of the most obvious corollary of moore's law is that progress is continuous; revolutionary directly implies non-continuous, and it seems unlikely that the development of feasible quantum computers would lead to a keeping of moore's law, rather than a breaking of it in one direction or the other.
      • How the hell did that get a +5 insightful? It's not in the least bit insightful... or rather, if it is, it's heavily nitpicking and silliness. Hmph. +3 would have been better. Mod me back down overrated, please.

    • Re:quantum entaglement (Score:5, Informative)

      by wass ( 72082 ) on Wednesday April 16, 2003 @02:32AM (#5742341)
      Nanotubes actually have significant potential for quantum computing. Nanotubes are much more than just a carbon 'wire', they are a well-structured crystal with a number of symmetry groups that can be exploited for interesting solid-state effects.

      I know some folks trying to make qubits out of nanotubes by patterning gates on them. Very very hard, they're so damn small standard lithography techniques are out the window.

      Nanotubes also have interesting phonon characteristics that make them good candidates for qubit systems. Also, it has been demonstrated that spin-orbit coupling in nanotubes can be drastically reduced, which can greatly enhance coherence times for spintronic qubits.

      So, if Intel or NASA is "only" looking at using these guys for interconnects, carbon nanotubes still have significant potential for revolutionary computing breakthroughs.

      • ...what they're doing here is far simpler than quantum computing. Basically, all they're doing is using nanotubes as conductive elements that have feature sizes smaller than what they can fab with copper (currently). To me, the interesting thing is that they have at least made an initial step toward fabbing the stuff.

        Also, as I recall, the major problem with using nanotubes in this way is going to be getting a number of them with similar characteristics. So far, no one's been able to get a good handle o

        • So far, no one's been able to get a good handle on how to really tailor properties finely (length, twist angle, etc).

          yeah, true. What we did when we measured tube resistance is put a drop of a solution containing a dispersion of nanotubes onto a small die-sized substrate. There was a trench etched in the substrate, with patterned metal leads, so by statistics, at the right tube concentration in solution, we'd get a few samples with one tube across the trench for each set of leads. thus, we can measure

    • Actually, bearing in mind that Moore's law discusses the number/density of transistors on a chip, the quantum computing probably won't "continue" it at all, where the nanotubes will. However, your point that *exponential growth in processing speed* may more likely be furthered by quantum conputing is well taken.

  • Sorry Dave (Score:5, Funny)

    by bcwalrus ( 514670 ) on Wednesday April 16, 2003 @01:45AM (#5742207)
    Now Mr. Bowman is supposed to pull Hal's nanotube? That's a bit hard.

  • need more info? (Score:5, Informative)

    by jayoyayo ( 650349 ) <(moc.u4qari) (ta) (nitsuj)> on Wednesday April 16, 2003 @01:50AM (#5742214)
    oddly enough, for more in depth information, check out [nasa.gov] the recorded answers they provide for integration into radio broadcasts.

  • I gotta do it.... (Score:4, Funny)

    by tiger_66_y2k ( 576892 ) on Wednesday April 16, 2003 @01:59AM (#5742240) Homepage
    Okay everybody you can LOOK at my new CPU, but what every you do, DON'T TAKE A PICTURE!!!

    *flash*
    **POP**

    ....shit....

    :)
  • What NASA has done is to make a switch from copper connectors to carbon nanotube connectors within the chip and (maybe) the boards. but while this IS a revolutionary step it will be a LONG while until we see this kind of chip for sale in teh home market because :
    • there are very few companies who are geared for this kind of manufacturing since everyone so far has been using copper for the past umpteen years
    • changing over to this kind of manufacturing will be a massive capital investment for a company, especially the companies in the East (asia not new york) where are a lot of these chips/boards are made
    • there are AFAIK no companies that make nanotubes in sufficient quantity and quality to feed the demand for the tubes at the moment
    • unless you are a gamer home computers are more than fast enough now for what we want (internet/email/minor word processing) this kind of tech will only benefit the "Power User" community..
    that said i should add that this is a pretty cool tech.. and i hope it works out.. after all .. <toolman>more POW-er urrhh urrh urrh</toolman>
    Suchetha

    • Re:This will take a while to seep down to home use (Score:5, Informative)

      by kinnell ( 607819 ) on Wednesday April 16, 2003 @02:39AM (#5742358)
      there are very few companies who are geared for this kind of manufacturing since everyone so far has been using copper for the past umpteen years

      That's completely untrue. For most of the history of the semiconductor industry, aluminium has been used, because the manufacturing process for copper was much more difficult. Copper has only recently become commonplace.

      changing over to this kind of manufacturing will be a massive capital investment for a company, especially the companies in the East (asia not new york) where are a lot of these chips/boards are made

      Changing to new manufacturing processes is a fact of life in the semiconductor industry and happens regularly. It always requires massive capital investment, yet somehow, they seem to manage (see above).

      there are AFAIK no companies that make nanotubes in sufficient quantity and quality to feed the demand for the tubes at the moment

      There are also no companies which manufacture nano scale copper wires for routing layers on ICs. This is because it's not done that way. Once you have a process for growing carbon nano-tubes on chips, you just have make it cost effective - just like any other semiconductor manufacturing technology.

      unless you are a gamer home computers are more than fast enough now for what we want (internet/email/minor word processing) this kind of tech will only benefit the "Power User" community..

      There's no amount of processing power that the desktop software industry will not be able to squander.

    • "only benefit the "Power User" community" (Score:5, Interesting)

      by da5idnetlimit.com ( 410908 ) on Wednesday April 16, 2003 @02:51AM (#5742388) Journal
      Ahem...

      People saying "unless you are a gamer home computers are more than fast enough now for what we want (internet/email/minor word processing) " are forgetting that ...

      1 - Starting Word 2024 will require 1.5 TeraFlops because every key you strike will require the calculation of two 8192 bytes key and the exchange of 1024 security tokens / sec, and we have to get ready to cope with that

      2 - My old and faithfull Dual PIII 1Ghz, that was once considered the fastest rig on my block is now just a piece of interesting junk that still allows me to play Quake and encode divxs at the same time, and LOTS of you just dream about doing it for real

      3 - it's not because i'm not a basic luser that immediatly jump categories and becomes a Power User. And if you think a softcore gamer or a hardcore Quaker is a "Power User", you never saw a real 16 CPU machine being "stability tested" for a round or ten of Quake @1024 fps, or the fastest Divx encode ever (11 minutes 8p)...

      4 - "internet/email/minor word processing" can be achieved since 486 DX2 66 with no problem and little fuss... I mean my mail Server/Firewall/Ftp/ Webserver/PDC is a Pentium 133 and it serves the need of 10 ppl...So stop complaining when we allow you the use of a 2 Ghz computer just so you can play Freecell @ 25 fps 8p

      Would be BOFH, hoping for Admin job...
    • unless you are a gamer home computers are more than fast enough now for what we want (internet/email/minor word processing) this kind of tech will only benefit the "Power User" community..

      False. Most improvements in processor design and fabrication have allowed processors to be made that are faster, smaller, and cheaper. Perhaps there aren't many users who care about faster, but everyone cares about smaller and cheaper.

    • unless you are a gamer home computers are more than fast enough now for what we want

      This won't stay true forever, though.

      My PII-400 is about 3 1/2 years old now, two generations old in Moore's Law terms, and it is indeed fast enough for most of what I do.

      However.

      If I try to play movie files using certain late-model codecs (MPEG-4 f'rinstance), I get jerky playback and poor A/V synchronization -- the processor just can't keep up with decoding the data streams in realtime.

      Software bloat shows no signs o

  • Moores law will never fail! (Score:4, Funny)

    by kinnell ( 607819 ) on Wednesday April 16, 2003 @02:16AM (#5742292)
    I for one am confident that the media and marketing people will be sufficiently creative to keep people believing in the Moores law myth well into the 23rd century.

  • Chimps... (Score:5, Funny)

    by K3lvin ( 624437 ) on Wednesday April 16, 2003 @02:16AM (#5742295)
    I read it "NASA Wires Chimps With Nanotubes"

    • I did too. I thought "Wow, NASA's back to animal experimentation in space." Followed by, "I wonder what PETA thinks about that."

      Oh well, back to the morning soda.

  • IBM pioneered Carbon Nanotubing (Score:5, Informative)

    by olePigeon (Wik) ( 661220 ) on Wednesday April 16, 2003 @02:54AM (#5742393)
    As reported in the April 27 (2001) issue of the journal Science, IBM researchers have built the world's first array of transistors out of carbon nanotubes -- tiny cylinders of carbon atoms that measure about 10 atoms across, are 500 times smaller than today's silicon-based transistors and are 1,000 times stronger than steel. The breakthrough bypasses the slow process of manipulating individual nanotubes one-by-one, and is more suitable for a future manufacturing process. Story is here [ibm.com].
    • This led to their creation of "pixie dust" which has enabled notebook hard drive capacities to rise. They found unique magnetic properties of "glass" when manipulating compounds on a molecular level.
    • I think you're missing the part where this technology is completely different from what IBM developed. The NASA research is referring to the interconnects between different parts of the IC, not the transistors themselves. So far as I know this is the first process that addresses interconnect technology with carbon nanotubes.

  • Nooooo! (Score:1, Funny)

    by Anonymous Coward
    We can't use nanotech! It'll be toxic and dangerous to breath in!

    WHATEVER YOU DO, DO NOT SNIFF YOUR CHIPS! :(

    So much for enjoying the new computer smell. :(
    • But what I heard from the herbal medicine industry is that grinding traditional medicine into nano meter size helps make it more potent.

  • One Million Amps!!!???? (Score:4, Funny)

    by addikt10 ( 461932 ) on Wednesday April 16, 2003 @03:52AM (#5742513)
    Understatement of the year:
    One advantage of using carbon nanotube interconnects within integrated circuits is that these interconnects have the ability to conduct very high currents, more than a million amperes of current in a one square centimeter area without any deterioration, which seems to be a problem with today's copper interconnects,"
    Dr Evil: One - Meelleeeooon Amps!!!!!

  • More info on their research on carbon tubes.. (Score:4, Informative)

    by cOdEgUru ( 181536 ) on Wednesday April 16, 2003 @04:03AM (#5742535) Homepage Journal
    Google points to here [nasatech.com]

    Also the interview mentions the fact that in October 2002, it was still in basic research form and could take as much as a couple of years to production and maybe a bit more for commercial purposes.

    But that still bodes well for us since Silicon will tide us through another 10 good years.

    Wish I werent 30 right now. The average lifespan looking like 70 (hopefully!) I just have 40 more years left....oopss.. Panic Attack!
  • More than a million Amps in a cm^2?? If space applications in the future are gonna need currents of a million amps going down a wire that feeds under your vertical bed, I sure as hell won't be an astronaut! :-)
  • These guys [liftport.com] are looking for 180,000 KM of the stuff, I wonder can the get it here.

    Oh and the need a big rock to tie it to as well....

  • I know where they got the idea from (Score:3, Funny)

    by Wicked L ( 663963 ) on Wednesday April 16, 2003 @06:01AM (#5742746)
    Just like the microwave, this is just yet another technological advancement made possible by Roswell.

  • That's all nice, but ... (Score:4, Informative)

    by DaneelGiskard ( 222145 ) on Wednesday April 16, 2003 @06:59AM (#5742908) Homepage
    ... processors are not the bottleneck in any way. They are already so fast that buses, caches and memory have a very hard time to keep up, not speeking about secondary or even tertiary memory at all. That's the real bottleneck these days, the buses to the caches and the caches/memory itself. Most of you know how many processor cycles are lost if some data cannot be pulled out of the cache, but must be pulled out of the memory or even the harddisk (we are speaking about millions of ns's here...).

    So I'd like to see some evolutionary/revolutionary inventions in these sectors, rather than making cpu's even faster and making the bottleneck of buses, caches and memories even larger...

    • Right now, the majority of space on chip is taken up by verious caches. A significant proportion of that space is taken up by wiring. Having much smaller wiring should allow much larger caches. A system with 8Mb on-chip cache (and a well-designed asynchronous algorythm for filling it) would hardly ever wait for the front-side-bus at all.

  • does it have to be carbon? (Score:3, Interesting)

    by crovira ( 10242 ) on Wednesday April 16, 2003 @07:31AM (#5743023) Homepage
    couldn't any atom in the valence group do as well? (I'm remembering my old chart of the elements and we could have silicon nanotubes too.)

    • Group IV elements (Score:4, Informative)

      by AlpineR ( 32307 ) <wagnerr@umich.edu> on Wednesday April 16, 2003 @08:16AM (#5743209) Homepage
      Good thought. The crystal structure of a silicon [jlab.org] wafer is the same as a carbon [jlab.org] diamond. Germanium [jlab.org], too, routinely grows in a diamond structure. But carbon also forms graphite, which is a sheetlike structure. Carbon nanotubes are essentially rolled up graphite sheets. But silicon and germanium are not stable in sheet structures, so they can't roll into nanotubes.

      However, intense research of carbon is what led to the discovery of buckyballs and nanotubes. Perhaps there other cool forms of silicon which are yet to be discovered.

      On a different topic, how do the NASA researchers propose to connect the nanotubes in a useful way? I can understand growing the tubes on a silicon wafer and filling in the surrounding space, but this just produces a bunch of parallel wires not a designed circuit.

      AlpineR

  • It seems to me that we could extend Moore's Law (observation) for another 7 years just by switching from copper and electron based bit transfer methods to fiber and light based bit transfer. Just like we went from copper telephone wires to fiber. Since electrons only travel at 1/10 the speed of light, we could theoretically have optical computers with FSB speeds of 8ghz! (quad pumped double data rate of course) That could hold us over until quantum computing arrives.

    • Since electrons only travel at 1/10 the speed of light

      ISTR that electricity in a wire actually travels by the transmission of momentum from one electron to another, not by the movement of electrons themselves, and that this transimission take place essentially at lightspeed. So I don't think the actual speed of the electrons is an issue.

  • Are they small enough? (Score:3, Interesting)

    by stevesliva ( 648202 ) on Wednesday April 16, 2003 @08:52AM (#5743427) Journal
    Yes, perhaps they promise less resistance than copper interconnect of the same size, but isn't a diameter of 100nm actually a bit large? Can nanotubes shrink, or is their diameter a chemical requirement? According to the International Technology Roadmap for Semiconductors [itrs.net], copper wiring pitch should now in 2003 already be 245nm. So with 50% spacing between those nanotubes, you're not even talking a 2x improvement in size over current interconnect. What if the things are too big to be used as interconnect for those 35nm gates we're supposed to see in 2007?

  • Ban It! (Score:4, Funny)

    by Ignorant Aardvark ( 632408 ) <cydeweys AT gmail DOT com> on Wednesday April 16, 2003 @09:13AM (#5743540) Homepage Journal
    I call for an immediate ban on all future use of nanotubes by NASA. I don't care about the "performance increases" they claim. All I care about is the health effects of nanotechnology - this must be banned before it gets out of control!

    -Crazy researcher from other recent /. article
  • SpaceDaily reports that NASA has come up with a way to wire microchips with nanotubes instead of copper interconnects.

    In other news, Intel's R&D department announced that mounting heatsink+fan on shuttles' thermal tiles can efficiently disspate heat during reentry into the Earth atmosphere.

  • NASA (Score:2, Funny)

    by JerryLs ( 587277 )
    Ok, this is the same NASA that only uses 486's in shuttles?

  • Not quite (Score:3, Informative)

    by dcmeserve ( 615081 ) on Wednesday April 16, 2003 @05:10PM (#5747305) Homepage Journal
    NASA has come up with a way to wire microchips with nanotubes instead of copper interconnects.

    If you read the article closely, you'll see it's not talking about about replacing all copper interconnect on the chip -- only a small portion, in fact: the vias. The carbon nanotube are being used only for the interconnect between metal layers, not between devices on the chip in general.

Slashdot Top Deals

If a 6600 used paper tape instead of core memory, it would use up tape at about 30 miles/second. -- Grishman, Assembly Language Programming

Close