Stories
Slash Boxes
Comments

SoylentNews is people

posted by n1 on Monday September 22 2014, @10:13PM   Printer-friendly
from the even-faster-in-turbo-mode dept.

Materials that change electrical states quickly could be capable of processing information 1000 times faster than silicon (Abstract), opening the doorway for developing computers that can perform calculations at well beyond current processing rates.

The present size and speed limitations of computer processors and memory could be overcome by replacing silicon with ‘phase-change materials’ (PCMs), which are capable of reversibly switching between two structural phases with different electrical states – one crystalline and conducting and the other glassy and insulating – in billionths of a second.

To improve the speed of computers, a solution is to increase the number of operations performed per device. Numerous operations in phase-change–based “in-memory” logic devices have previously been achieved using crystallization, but they show slow speeds, mostly due to a trade-off between the crystallization speed and stability of the initialized-glassy states. Here, we instead control melting processes to perform logic operations. Ultrafast melting speeds and diverse operations were achieved.

Multiple operations in phase-change–based logic devices have been achieved using crystallization; however, they can achieve mostly speeds of several hundreds of nanoseconds. A difficulty also arises from the trade-off between the speed of crystallization and long-term stability of the amorphous phase. We here instead control the process of melting through premelting disordering effects, while maintaining the superior advantage of phase-change–based logic devices over silicon-based logic devices. A melting speed of just 900 ps was achieved to perform multiple Boolean algebraic operations (e.g., NOR and NOT). Ab initio molecular-dynamics simulations and in situ electrical characterization revealed the origin (i.e., bond buckling of atoms) and kinetics (e.g., discontinuouslike behavior) of melting through premelting disordering, which were key to increasing the melting speeds. By a subtle investigation of the well-characterized phase-transition behavior, this simple method provides an elegant solution to boost significantly the speed of phase-change–based in-memory logic devices, thus paving the way for achieving computers that can perform computations approaching terahertz processing rates.

This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 2) by opinionated_science on Monday September 22 2014, @10:36PM

    by opinionated_science (4031) on Monday September 22 2014, @10:36PM (#96951)

    If you read the article that is cited in the press release, they demonstrate the system at 875K.

    In practice this means a CPU/Memory of this could be kept hot, and have it as a working condition. By "pre-heating" the melting takes less time (10ps).

    I'd like to see it as memory first, so we can have single cycle modern CPUs!!!

    • (Score: 0) by Anonymous Coward on Tuesday September 23 2014, @12:28AM

      by Anonymous Coward on Tuesday September 23 2014, @12:28AM (#96976)

      What sort of energy do you need to keep it at that temp. That sounds expensivy....

      • (Score: 2) by EvilJim on Tuesday September 23 2014, @04:02AM

        by EvilJim (2501) on Tuesday September 23 2014, @04:02AM (#97028) Journal

        stick a uranium core in the centre of the IC's :)

    • (Score: 2) by c0lo on Tuesday September 23 2014, @02:28AM

      by c0lo (156) Subscriber Badge on Tuesday September 23 2014, @02:28AM (#97005) Journal

      Multiple operations in phase-change–based logic devices have been achieved using crystallization; however, they can achieve mostly speeds of several hundreds of nanoseconds. A difficulty also arises from the trade-off between the speed of crystallization and long-term stability of the amorphous phase.

      So, melting is fast (by pre-melting). How about toggling the bit back, assumed as requiring crystallization? What good would it be a memory/CPU that stores/processes the zeroes in 1 nanosecond but the ones take hundred times longer?

      --
      https://www.youtube.com/watch?v=aoFiw2jMy-0 https://soylentnews.org/~MichaelDavidCrawford
      • (Score: 2) by EvilJim on Tuesday September 23 2014, @04:00AM

        by EvilJim (2501) on Tuesday September 23 2014, @04:00AM (#97026) Journal

        A quick burst from the fire extinguisher would probably resolve that... and all the other supposed-to-be-melted bits

  • (Score: 2) by takyon on Tuesday September 23 2014, @12:20AM

    by takyon (881) <takyonNO@SPAMsoylentnews.org> on Tuesday September 23 2014, @12:20AM (#96971) Journal

    Optalysys [hpcwire.com]

    A 340 gigaflops proof-of-concept model is slated for launch in January 2015, sufficient to analyze large data sets, and produce complex model simulations in a laboratory environment, according to the company.

    The company is developing two products: a ‘Big Data’ analysis system and an Optical Solver Supercomputer, both on track for a 2017 launch.

    The analysis unit works in tandem with a traditional supercomputer. Initial models will start at 1.32 petaflops and will ramp up to 300 petaflops by 2020.

    The Optalysys Optical Solver Supercomputer will initially offer 9 petaflops of compute power, increasing to 17.1 exaflops by 2020.

    Perhaps the most impressive trait of all is the reduced energy footprint. Power remains one of the foremost barriers to reaching exascale with a traditional silicon processor approach, but these optical computers are said to need only a standard mains supply. Estimated running cost: just £2,100 per year (US$3,500).

    To compare, scaling up today’s technology to exascale levels would require at least 200MW of power, and the current fastest supercomputer, Tianhe-2 in Guangzhou, China, requires 24 MW per year (including cooling) at a cost of about $21 million per year.

    Optalysys Ltd. raised over £400,000 (US$675,000) in seed money earlier this year, which enabled it to bring its innovative technology to NASA Technology Readiness Level 4 ahead of schedule.

    --
    [SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
  • (Score: 0) by Anonymous Coward on Tuesday September 23 2014, @03:21AM

    by Anonymous Coward on Tuesday September 23 2014, @03:21AM (#97019)

    Either this technology won't pan out properly or it DOES and the NSA/GCHQ force the inventors to sit on it and the test results in the name of 'national/(global?) security'.

    Faster computers means faster, stronger crypto--pure and simple. Can't have that available to the public at large in the midst of the ongoing 'War On Terror'.

    The Feds get/create top-of-the-line computer gear for themselves and the consumers get the 'watered down' retail equivalents as the PC hardware vendors are likely pressured by NSA, et. al. to hold back computer innovation in the name of 'national/(global?) security' for the aformentioned reason: Faster computers means faster, stronger crypto--pure and simple.