2012-02-28

Co.Exist

The Unbreakable Smartphone That Lasts For Weeks Without Recharging

It sounds utopian, but a new kind of memory is going to make your phone drastically more efficient and less prone to breaking down.

Stuart Parkin’s digital storage research is part of the reason why the video stored on your cell phone works. It has also contributed to Google and Facebook’s ability to build giant data centers. The IBM researcher’s work on data storage has changed the way we use electronic devices. Now he’s about to do it again--this time, by soaring past Moore’s Law with a new kind of memory technology that’s 100 times faster and far more energy efficient than what we have now.

The technology is called racetrack memory, and eight or so years from now, it might be what’s storing data in your laptop or cell phone. Racetrack memory is more stable than flash (which is what Apple is rumored to be using for memory in the next generation of MacBooks), allows for long battery life, and stores unfathomable amounts of data.

Imagine: a nearly unbreakable smartphone that can store thousands of movies and lasts for weeks without needing to be recharged. In one fell swoop, racetrack memory could mitigate the problem of pricey devices constantly needing to be replaced and alleviate pressure on the power grid.

If that happens, we might have Parkin’s work on spintronics--a technology that uses an electron’s spin and "magnetic moment" in addition to its electric charge--to thank. In the case of magnetic memory, currents of spin-polarized electrons move magnetic data up and down a vertical racetrack (hence the name) on a silicon chip at a speed of hundreds of miles per hour. It’s this move into the third dimension (up and down) that allows racetrack memory to bypass Moore’s Law.

Ultimately, IBM hopes that racetrack memory devices can be wiped and rewritten millions of times. Flash drives often lose reliability after a bit undergoes 100,000 rewrites. It stand to reason, then, that devices using racetrack memory should last significantly longer than their flash counterparts.

Stuart in his lab.

The concept for racetrack memory has been around since 2008, but it was only in December 2011 that the IBM team unveiled the first fully functioning prototype on a single chip.

There is a ways to go before racetrack memory is commercialized, but in the meantime, we can look forward to seeing magnetic random access memory (MRAM)--a chip that operates using magnets instead of electricity--in electronics within the next two to three years. MRAM isn’t quite as worldchanging as racetrack memory, but it will mean that smartphones can last for days longer than they do now without needing to be charged.

Add New Comment

3 Comments

  • Nobody

    Completely agree with Ajent J.  This article is simply silly.  For one thing, is memory what uses up energy on your smartphone?  Its not, so having more efficient memory will not really change how long you can go between charges.  Processor energy use is way more important.

  • Gary

    That pic with Sturat in his lab has a phone that certainty looks unbreakable.  I don't want to see him try to put it in his pocket though.

    If you look in the lower right corner and you can see where it plugs into the wall.  Just like my old Samsung,  amazing.

    !00 times faster and less energy in 5-10 years by my math does not seem to break Moore's law but might be just the thing then to keep Moore's law going.

  • Ajent J

    This is my first visit to co.exist and I am very
    disappointed. As an expert in this field I feel compelled to point out to the
    other readers that this article is a biased and uninformed glimpse into the
    world of data storage and emerging memories, specifically magnetoresistive
    memories. Ms. Schwartz’ gushes praise and techno-hype as if she was an agent
    for Dr. Parkin or a PR person for IBM. A human-interest story on Dr. Parkin
    could have been nice; he is an interesting and brilliant man. A short,
    well-informed, technical story about the promise of emerging memories could
    have been enlightening.  But the broad
    statements crediting Dr. Parkin for all progress in data storage is insulting
    to the hundreds of other dedicated scientists who contributed as much or more,
    and the lack of any perspective about the worldwide efforts to commercialize
    new data storage technologies leaves me wondering why this was written and why
    it was chosen to appear here.