back to article Boffins flip the unflippable: Meet the latest storage extender contender

A team of international researchers from the University of Nottingham has found a new material – copper manganese arsenide – which could be the future of computer storage, according to a study published in Science. Using a semiconductor as a template, a thin film of CuMnAs(copper manganese arsenide) crystals was grown on the …

  1. Pascal Monett Silver badge
    Windows

    "it'll take years"

    No, at this point it will take decades.

    Spintronics ? Purely lab-based. I'm impressed that they've found another way to store digital information. I'm happy that an EM blast cannot erase it. I am however too old to be taken in by the hype.

    Maybe, just maybe, a copper-whatever-arsenide SDD thingy will be commercially available before I die.

    I won't hold my breath. Not until I start seeing those wonderful nano-carbon-tube batteries I've already been promised a decade ago.

    1. allthecoolshortnamesweretaken

      Re: "it'll take years"

      You'll be using a spintronic-based tablet powered by nano-carbon-tube batteries to watch the inflight movie on the space elevator...

      Btw, I'm a bit confused. Isn't Moore's law about CPUs, not storage? Did I get this wrong?

      1. PleebSmasher
        Dead Vulture

        Re: "it'll take years"

        "Btw, I'm a bit confused. Isn't Moore's law about CPUs, not storage? Did I get this wrong?"

        Moore's law is about whatever the author wants it to be about.

      2. Mage Silver badge

        Re: "it'll take years"

        It was about doubling of transistors in ANY chip, originally same size chip and annually, then 18 months, 2 years, any size chip or anything!

        It's been a bit dead since about 2003. When CPUs (or anything) used 90nm, that was the general geometry. But 14nm chips are not really as the 14nm isn't even close to average size, but smallest feature.

        Since this might be more like bubble memory rather than spinning disk, I guess it's like a chip?

        Bubble memory WAS used in production, a great future claimed in 1970s, but never survived due to Static Ram + Lithium cell, then Flash and also miniature high capacity HDD and even MO 128M and larger drives killed it.

        It could be a decade or never before we see this.

        1. Jeffrey Nonken Silver badge

          Re: "it'll take years"

          Wow, somebody who remembers bubble memory. ... Well, somebody else, I mean. I gathered at the time that the real reason it never took flight was they never really got it working reliably. But it's quite possible it was due to, as you say, other technologies overtaking it.

          But... bubble memory? No, think core memory! Stored as individual magnetic cells, programmed via magnetic pulse, read back by magnetic pulse, persistent without power applied... at least they've apparently managed not to have the read function be destructive. Took longer to read than write core memory due to needing to write back what it read.

          I wonder if it uses a sense wire?

    2. petewadley

      Re: "it'll take years"

      Spintronics is not purely lab-based. The read heads on your hard disk are entirely spintronic in functionality and have enabled the age of mass data storage we are experiencing.

      I share your scepticism and it is early stage, however, it is a completely new paradigm in memory storage and I believe will fill some commercial role in the not distant future because of its unique attributes.

      we will have to see.....

      1. DougS Silver badge

        Moore's Law

        No, Moore's Law lived well past 2003, arguably it is still going but at least continued until this year. Moore's Law is not about the feature size, it is about doubling the number of transistors on a chip every couple years. That has still been happening. Intel is changing their cadence to every three years starting this year, but TSMC has recently accelerated its timeline to catch up and possibly pass Intel - if they really do they might keep Moore's Law alive until 2020 or so.

        It isn't that we've hit a technological wall, the wall is really more economic at this point. Fewer and fewer companies can justify the expense of designing leading edge chips so there are fewer potential customers who need higher and higher volumes to make it work.

    3. Katie Saucey
      Meh

      Re: "it'll take years"

      Yep, about five years. I look forward to picking my new copper-whatever-arsenic HDD up in my hydrogen powered flying car, also just around the corner, in 5 short years.

  2. allthecoolshortnamesweretaken

    "Isn't there a XKCD cartoon about such things?"

    "Yes."

    1. Mage Silver badge
      Coat

      XKCD

      He's got one on his desk. So is it years, decades or "I quite like having the only hover car"

      1. petewadley

        Re: XKCD

        I do have one on my desk but it is a single-bit proof-of-concept device. Pretty astonishing considering the research was published in March, but a far cry from being a usable device. Competing with existing and established technologies takes a huge amount of materials research. Give us a chance....

  3. Mage Silver badge
    Boffin

    manipulating the magnetic ordering of antiferromagnets is tricky.

    If not nearly impossible. However makes what you manage to write quite robust?

    1. Anonymous Coward
      Anonymous Coward

      Re: manipulating the magnetic ordering of antiferromagnets is tricky.

      They manipulate a pair instead of a singleton. That 's probably the operative difference that makes it work. That's from the PhysOrg article.

      http://phys.org/news/2016-05-spintronics-material-huge-storage.html?utm_source=nwletter&utm_medium=email&utm_campaign=daily-nwletter

  4. anthonyhegedus Silver badge

    ummmm

    I don't understand this properly. Is this way-off-in-the-future storage technology mechanical? By mechanical, I mean will it work like a hard disk but just data being packed more tightly?

  5. Michael H.F. Wilkinson Silver badge
    Coat

    If you have a PhD in spintronics...

    does that make you a spin doctor?

    Sorry, couldn't resist.

    OK, time to go. My hat (genuine Panama) and coat please

    1. Anonymous Coward
      Anonymous Coward

      Re: If you have a PhD in spintronics...

      " My hat (genuine Panama)..."

      The one with the €500 notes tucked in the hat band?

  6. Rol Silver badge
    Coat

    Timely storage

    Save all your data to a 2TB disk and then tomorrow delete it and then save a new slew of data.

    Keep repeating this until the disk eventually dies, which by my estimate would suggest quantribillions of GB's would have gone through this process.

    When they invent time travel plug in your 17.1 USB time portal and access the drive by date as well as block, thus giving you access to quantribillions of Gig's of data, with no loss of integrity.

    I best fly to Delaware now before the time trolls get in first. Where's my coat?

  7. Dave 32

    Hematite

    So, is anyone playing with the magnetic properties of Hematite?

    https://en.wikipedia.org/wiki/Hematite#Magnetism

    Dave

  8. hellwig Silver badge

    Isn't Size an Issue?

    I thought the problem with magnetic recording was that the size of the write head could only be made so small. So, are they able to make the write head small enough to take advantage of the smaller footprint of these anti-ferro-magnetic crystals (understanding that it is a different technology and uses different read/write heads)?

  9. Sgt_Oddball Silver badge
    Coat

    CuMnAs?

    The main compound that makes this magic work is CuM'n'As? Surely you can't be serious?

    Mines the one with page 3 of the guardian stuffed inside the pockets...

    1. Nixinkome

      Re: CuMnAs?

      81 102 Periodic Table Code!

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Biting the hand that feeds IT © 1998–2019