Slashdot Mirror

← Back to Stories (view on slashdot.org)

Biological Supercomputers Powered By ATP Could Be A Reality Some Day (dispatchtribunal.com)

Posted by BeauHD on from the molecular-motor-propelled-agents-in-nanofabricated-networks dept.

hypnosec writes: Our cells are powered by Adenosine triphosphate (ATP) and according to a new study, they could be a power source for the next generation of biological supercomputers capable of processing information very quickly and accurately using parallel networks in the same way that massive electronic super computers do. Published in the Proceedings of the National Academy of Sciences (PNAS), the paper describes a model of biological computer that is effectively a very complex network in a very small area, and is based on a combination of geometrical modeling and engineering know-how (on the nano scale). Researchers involved with the study claim that it is the first step in showing that this kind of biological supercomputer can actually work.

23 of 66 comments (clear)

  1. Old news by penguinoid · · Score: 4, Funny

    I'm a biological supercomputer that runs on ATP. Well, sugar and caffeine really, but it's converted to ATP in between.

    --
    Don't waste your vote! Vote for whoever you want, unless you live in a swing state it won't matter anyways
    1. Re:Old news by Anonymous Coward · · Score: 1

      I'm greener than you. I run on an ethanol blend.

      So there.

    2. Re:Old news by penguinoid · · Score: 4, Funny

      In case you're wondering, use a hardware-based neural net to learn and store data. All my components are fully biodegradable and non-toxic. My capabilities include real-time video processing and object identification, full voice recognition, and the ability to pass the Turing Test. You can download both my blueprints and full base code here. The full download is pretty small; smaller than a modern computer game in fact.

      --
      Don't waste your vote! Vote for whoever you want, unless you live in a swing state it won't matter anyways
    3. Re:Old news by codeButcher · · Score: 1

      In soviet Russia, the Matrix exists in you.

      --
      Free, as in your money being freed from the confines of your account.
    4. Re:Old news by MaxSmoke · · Score: 2

      I bet you release a toxic cloud time to time..

    5. Re:Old news by hene · · Score: 3, Funny
      Once they get this production ready, I hope that bug fixes are compatible because I have a long list of those.
      • Algorithms that background process(es) use to defragmentation and dedublicating data are severely flawed and continually cause data loss.
      • The storage medium is ridiculously unstable and corrupts data in unacceptable rate.
      • When retrieving from the cold storage, latency is too high to be practical in almost any real world situation.

      Just to mention a few. Could use that quicker information processing they advertise too.

    6. Re:Old news by JoeMerchant · · Score: 2

      The full download is pretty small; smaller than a modern computer game in fact.

      That's just the wetware design, you forgot to mention the 30+ years of programming effort it took to get you up to being a reasonably functional member of the collective.

    7. Re:Old news by Impy+the+Impiuos+Imp · · Score: 1

      Garbage in shit out

      --
      (-1: Post disagrees with my already-settled worldview) is not a valid mod option.
    8. Re:Old news by Solandri · · Score: 1

      I wouldn't recommend this. I've tried using it for a while and it's very finicky. For one, it needs to be maintained within a tight temperature range of between about 35-42 C. Allow it to drift outside that range even for a few minutes and you're looking at total loss of the hardware and neural net. It has some built-in temperature regulating capability, but still needs supplemental insulation and heating at cooler temperatures, and active cooling at higher temps. Blocking the air vent results in system hibernation in less than a minute, and total loss of the system within 5-10 minutes.

      Also, you don't immediately get a fully functional system. It takes a long time - a couple decades - before it reaches stability. Granted most of that time is spent priming the neural net. But the neural net is slow too. It seems to function based on creating new physical connections which take time to grow, instead of virtual ones which can be made or destroyed instantly. In contrast, a silicon-based system can be up and running in a few months, with priming the neural net taking about a year, quicker if you can dump in the data from a previous iteration (also something you can't do with these biological systems).

      It does have some nifty self-repair mechanisms though, allowing it to last many decades with little to no maintenance. The silicon-based hardware I've tried only lasts a few years to a decade, with the longest-lived example only making it to 2 decades because it was barely used. But the ability to quickly copy data from previous hardware makes this a moot point.

      The biological neural net has some strange quirks too. After about a decade and half of priming, it starts spending more of its time trying to spread its blueprints and base code around, instead of solving the problems I assign it. I mean sure, when it succeeds it results in more neural nets which increases the overall computational capability of my systems. But as I said, each one takes over a decade to prime, so that represents a lot of time and resources you have to dump into a new system before you see any return on investment.

  2. Re:The problem by mysidia · · Score: 1

    Perhaps we're 3 years away from an even more complete lab demo, but better and better lab demos are aw way of eventually amounting to something too.

    Is it the new thing on slashdot to criticize developments that are 'lab demos' ? I just noted This the other day, on the new research breakthrough in super batteries.

    Perpetually 10 years away, along with the solar panel to fill it. We'll have year of the Linux Desktop about 9 years before you can buy one of these batteries on eBay

  3. Re:The problem by fuzzyfuzzyfungus · · Score: 2

    What's really sneaky is that every lab demo of a biological supercomputer is actually a production-ready biological supercomputer showing off some crude little toy that will be 5 years out for the next few decades.

    The trick is just to ignore whatever is on the benchtop and grab the guy talking about it.

  4. Not no, but hell no! by Hartree · · Score: 1

    My office mates are already bad enough at stealing my Doritos without automating it!

    1. Re:Not no, but hell no! by DigiShaman · · Score: 1

      "Resistance is futile! We will assimilate the Doritos and its biological and technological distinctiveness to our own."

      --
      Life is not for the lazy.
  5. PNAS? by Stoutlimb · · Score: 1

    If I worked there, how would I tell people?

    1. Re:PNAS? by wbr1 · · Score: 2

      We had some female visitors to the big PSAS building today....

      --
      Silence is a state of mime.
  6. ATP source? by Anonymous Coward · · Score: 2, Funny

    so finally humans are going to be milked to produce ATP.. where have I seen this in the movies.. matrix?

  7. Computer viruses could someday be real viruses by BlueCoder · · Score: 1

    It also means a biological computer could catch a cold.

    Talk about hacking...

  8. Re:Seems useless and impractical by xtal · · Score: 1

    The most complicated and efficient computing device in the known universe is the human brain. Nothing else comes even remotely close.

    Engineering organs also provides the option to well, engineer brains.

    Interesting times.

    --
    ..don't panic
  9. Did I miss something? by Dcnjoe60 · · Score: 2

    Our cells are powered by Adenosine triphosphate (ATP) and according to a new study, they could be a power source for the next generation of biological supercomputers

    When did we get this generation of biological supercomputers?

  10. Re:The problem by Dcnjoe60 · · Score: 1

    Is it the new thing on slashdot to criticize developments that are 'lab demos' ?

    No, we've been critizing lab demos for years. Could be because most lab demos never make it past that stage and will always remain a novelty. Kind of like you can power a clock from a potato, but it doesn't scale very well as an energy source.

  11. Re:The problem by KGIII · · Score: 1

    I have not crunched any numbers but I bet the folks who say this are right more often than not. How often do lab demos turn into viable commercial products?

    --
    "So long and thanks for all the fish."
  12. Mentats by robinsonne · · Score: 1

    I for one welcome our future biologic, ATP-powered Mentat overlords.

  13. Neal Stephenson by The-Ixian · · Score: 1

    Anyone else think of that underwater orgy thing from Diamond Age when you read this summary?

    --
    My eyes reflect the stars and a smile lights up my face.