New microchip uses 30 times less power

New microchip uses 30 times less power

Summary: Scientists at Rice University have created a microchip that uses 30 times less electricity while running seven times faster than today's best technology.The U.

SHARE:

Scientists at Rice University have created a microchip that uses 30 times less electricity while running seven times faster than today's best technology.

The U.S.-Singapore team developing the technology, named PCMOS, revealed results at the International Solid-State Circuits Conference (ISSCC), the world's premier forum for engineers working at the cutting edge of integrated-circuit design.

Conceived by Rice University professor Krishna Palem, PCMOS piggybacks on the "complementary metal-oxide semiconductor" technology, or CMOS, that chipmakers already use. This means chipmakers won't have to buy new equipment to support PCMOS, or "probabilistic" CMOS. Although PCMOS runs on standard silicon, it breaks with computing's past by abandoning the Boolean logic that has (thus far) been used in all digital computers. PCMOS instead uses probabilistic logic, a new form of logic developed by Palem and his doctoral student, Lakshmi Chakrapani.

"A significant achievement here is the validation of Rice's probabilistic analogue to Boolean logic using PCMOS," said Shekhar Borkar, an Intel Fellow and director of Intel's Microprocessor Technology Lab. "Coupled with the significant energy and speed advantages that PCMOS offers, this logic will prove extremely important because basic physics dictates that future transistor-based logic will need probabilistic methods."

According to the researchers, silicon transistors become increasingly 'noisy' as they get smaller, but engineers have historically dealt with this by boosting the operating voltage to overpower the noise and ensure accurate calculations. Chips with more and smaller transistors are consequently more power-hungry, they said.

"PCMOS is fundamentally different," Palem said. "We lower the voltage dramatically and deal with the resulting computational errors by embracing the errors and uncertainties through probabilistic logic."

PCMOS was jointly validated by Rice and Nanyang Technological University in Singapore via a joint institute that Palem founded in 2007, the Institute for Sustainable Nanoelectronics.

(from "The Explicit Use of Probability in CMOS Designs and the ITRS Roadmap: From Ultra-low Energy Computing to a Probabilistic Era of Moore's Law for CMOS," Sept. 2005)

The prototypes were application-specific integrated circuits that were designed solely for encryption. Unlike the general-purpose microprocessors that power PCs and laptops, ASICs are designed for a specific purpose, and they are "embedded" by the millions each year in a growing constellation of products like automobiles, cell phones, MRI scanners and electronic toys.

The Rice-NTU team plans to follow its proof-of-concept work on encryption with proof-of-concept tests on microchips for cell phones, graphics cards and medical implants.

Palem said PCMOS is ideally suited for encryption, a process that relies on generating random numbers. It's equally well-suited for graphics, the scientists say, for different reasons:

In a streaming video application on a cell phone, for example, it is unnecessary to conduct precise calculations. The small screen, combined with the human brain's ability to process less-than-perfect pictures, results in a case where the picture looks just as good with a calculation that's only approximately correct.

If PCMOS can slash energy use for embedded ASICs in key devices, the implications are enormous. For consumers, it could mean the difference between charging a cell phone every few weeks instead of every few days. Globally, that mean a smaller carbon footprint for the IT industry.

Palem said he hopes PCMOS technology will enter the embedded computing market in as little as four years. Palem's PCMOS research was funded by the Defense Advanced Research Projects Agency and Intel Corp.

Topics: Mobility, Hardware, Processors

Andrew Nusca

About Andrew Nusca

Andrew Nusca is a former writer-editor for ZDNet and contributor to CNET. During his tenure, he was the editor of SmartPlanet, ZDNet's sister site about innovation.

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.

Talkback

17 comments
Log in or register to join the discussion
  • good luck

    analog calculators have been arround for some time.
    I don't see this technology ending up on servers since it is not 100% exact.
    As for the cell phone, the power hog is the transmitter and LCD.
    Linux Geek
  • RE: New microchip uses 30 times less power

    Microsoft will find a way to make it run slow.
    kozmcrae
    • With probablistic BSOD's...

      Oh wait, that already comes standard...

      (I'm joking, it's not *that* bad)
      T1Oracle
  • 30 times less?

    So if a given processor uses 1 watt, this one will use 30 watts less?

    How about 1/30th?
    Eriamjh
    • Yes, 30 times les means...

      ...you can buy seven of them and run an electric bicycle with the energy you make from them.

      What an interesting concept.
      arcebus@...
    • Impossible

      A silicon transistor requires a 0.7v bias voltage to work. This means that you need MORE voltage than that to "switch" the transistor. Germanium only requires a 0.3v bias - which means that it can operate at lower voltages. I'm not sure about GaAs (Gallium Arsenide).

      They must be talking about CURRENT and not voltage. But 30 times less current doesn't sound realistic (using existing methods).
      Roger Ramjet
      • Terms like current become obsolete

        when you're talking about a handful of, or even single electrons.

        Similarly voltage, although I dont know the mechanism they are using. I'd suspect they are using the probability of an electron being in a particular place at a particular time to do the work...

        Processors and other low-voltage integrated circuits that could use this technology typically operate at 3.3v overall; there are millions of transistors, and using standard gating techniques you are quite right, there wont be enough electrons available to saturate the junctions - but these new things dont work like that.

        I'm off to do some research before I have a clue how they do tho. :)
        SiO2
    • Ditto

      I, too, find it strange when people use "times" to describe LESS of something. It seems to fail from a logic standpoint.

      If a given processor uses 30 watts, would a processor that uses 30x less power use 1/30th of that (i.e., 1 watt)?
      ParrotHeadFL
  • RE: New microchip uses 30 times less power

    No such thing. Perhaps 1/30 (one-thirtieth) the power.
    nrkmann@...
  • I'm somewhat confused...

    Rather than actually being designed to need less power, aren't they talking about simply dealing with the error rates of using less power with some sort of statistical analysis?
    D. W. Bierbaum
  • Buzz Words

    This article is just like the ones you see every few years about flying cars or new ways to construct solar cells. Lots and lots of hype, no real information or data, and certainly nothing that is even close to being put into production.

    For example:

    "30 times less electricity": No units, just an expression that means nothing

    "follow proof-of-concept work on encryption with proof-of-concept test": That is, this is a theory which has been looked at . . . theoretically

    "If PCMOS can slash energy use": If, if, if

    "Hopes PCMOS technology will enter the embedded computing market in as little as four years": FOUR YEARS!?
    LadyGray
    • have you really check this out, or or are you just angry?

      Flying cars are a cool, fun, idea that is easy to get a bit over excited about, but for reasons I won't get in toto, the things that are available now, well are at best an expensive niche product. Solar tech continues to improve/become more and more cost effective>where's the parallel? My time would have probbably been better spent checking out the universitty studdy link etc., but once in a while the lack of logic that is omnipresent in human discourse, and the drive to share poorly thought-out opinions gets my attention.
      w
      burt&theband
    • my msg went out before I was finished...

      You do have a point wwith the "30 times less energy" thing, sensasional headlines that are never followed-up on get my goat too, but as a glimpse at a promising idea this was an acceptable article, and one is always free to dig deeper for more tech details. (I have to laugh at myself for wasting valuable time commenting on a bad comment to an average qualitty info-glimpse article(hope I don't sound as angry as l.g.)) btw-please excuse the double post and typos, but I'm blind and my screen-reader doesn't work well in on-line edditing situations>almost never wast my time on this type of venting/"the facts m'am, just the facts"....

      burt&theband
      • I'd like to point out...

        ...that I've provided links to the press release from Rice as well as several papers by the author of the study detailing the technology in question.

        andrew.nusca
        • I apologize

          I didn't mean to offend, so much as point out some problems so obvious that I assumed my post would just be ignored. I have read the technical paper, which was a theoretical paper (the authors didn't do any basic hands-on research), but was intended to only be a theoretical paper which could be built on in later years (though I don't see any further papers on the subject). I have gone to the website of the ISNE (which is still being updated), and find mostly stock photographs of circuits.

          It is an interesting theory, which may be proven to have some merit. However, when circuits become this small and specialized, there are problems that creep in, despite all the theories of how it should work. Personally, I would prefer to see news articles that detail some of the data derived from physical working models. Looking at graphs of computer simulations and models doesn't produce the same level of satisfaction.

          My comment about solar cells deals with the many attempts through the past decades (since the 1950's, at least) to produce such cells either more cheaply or better. If they were produced more cheaply, or with massively greater efficiency, we would all be using them.

          And so time will tell with this particular bit of technology. In twenty years, it may be what keeps laptops running for weeks instead of hours.

          However, the comments about my original post show me that I should be more careful on my phrasing of such things. In keeping my comments somewhat short, I may have actually cut out what I intended to convey.
          LadyGray
  • Uses 30 times less power?

    Obviously these guys are all scientists and not English majors. It should be one-thirtieth the power. ?Uses thirty times less power? is an oxymoron. It?s like saying ?trustworthy bankers?.
    Hates Idiots
  • Cell Phones?

    Please correct me if I'm wrong, but I believe that most of the power required with wireless devices (like cell phones) is due to the transmitter, not the processor.
    Isn't that why cell phones last so much longer when you aren't using them to make calls?
    justincase@...