Computers have speed limit as unbreakable as speed of light, say physicists

Computers have speed limit as unbreakable as speed of light, say physicists

Summary: A pair of physicists have shown that if processors continue to accelerate in accordance to Moore's Law, we'll hit the wall of faster processing in roughly 75 years.The curtain will eventually come down for silicon in today's manufacturing methods once engineers can no longer further shrink transistors and the copper wires that connect them.

TOPICS: Hardware, CXO

A pair of physicists have shown that if processors continue to accelerate in accordance to Moore's Law, we'll hit the wall of faster processing in roughly 75 years.

The curtain will eventually come down for silicon in today's manufacturing methods once engineers can no longer further shrink transistors and the copper wires that connect them. Processor fabrication using new technologies such as imprint lithography, graphene, and quantum computing will continue to yield faster and smaller chips. Nonetheless, those advanced techniques only stave off the absolute ceiling for speed, no matter how small the components get, according to professors Lev Levitin and Tommaso Toffoli at Boston University's Department of Electrical & Computer Engineering.  As Inside Science reports, the two have slapped a speed limit on computing.

A UCSB researcher holds a mask that will be used in building a 4 qbit chip (Credit: James Neeley via Flickr)

In a recently published paper in the journal Physical Review Letters, Levitin and Toffoli present an equation for the minimum sliver of time it takes for a single computation to occur and establishes the speed limit for all possible computers.

Using their equation, Levitin and Toffoli calculated that, for every unit of energy, a perfect quantum computer spits out ten quadrillion more operations each second than today's fastest processors, according to Inside Science.

"If we believe in Moore's law ... then it would take about 75 to 80 years to achieve this quantum limit," Levitin said. ??"No system can overcome that limit. It doesn't depend on the physical nature of the system or how it's implemented, what algorithm you use for computation … any choice of hardware and software. This bound poses an absolute law of nature, just like the speed of light."

The physicists point out that technological barriers might slow down Moore's law as we approach this limit.

Scott Aaronson, a pioneer in quantum computing and assistant professor of electrical engineering and computer science at MIT believes that the forecast for 75 years is optimistic.?? Moore's Law, he said, probably won't hold for more than 20 years. ??That may be so, but predictions made over the last few decades about the death of Moore's Law have been wrong.

Here's another quote from Aaronson:  ??"From a theorist's perspective, it's good to know that fundamental limits are there, sort of an absolute ceiling. You may say it's disappointing that we can't build infinitely fast computers, but as a picture of the world, if you have a theory of physics allows for ?infinitely fast computation, there could be a problem with that theory."

In your view, how much longer till the demise of Moore's Law?  Post in Talkback.

Topics: Hardware, CXO

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


Log in or register to join the discussion
  • Facinating

    I always knew there would be a wall somewhere, but what will be needed with that speed.

    But then aggain, this study uses current day technologies and materials, maybie in 70 years we will have some kind of new technology to make us bypass this wall... but then a new wall will present itself.
    • No, not "current day technologies and materials" it's an absolute limit

      Just as no technology will be able to break the speed of light limit no technology will be able to break this amximum computation limit.

      <i>" equation for the minimum sliver of time it takes for a single computation to occur and establishes the speed limit for all possible computers"</i>

      That limit is derived from first principles, it has nothing to do with the technology used.
      The Mentalist
      • Ummm

        I said bypass this wall, not break this wall. It's possible that in 70 we find a way to use either a new material or a new technology or even a new method of thinking to bypass the wall in a way to improve the computationnal speed of a computer.

        That does not mean that this study is wrong. It's just that it might be possible to use something to bypass a limitation(like every single limitation in existance).

        Hell it could be that the way to bypass the limitation is a new way to optimise the code to "improve" the processing speed yet the true speed is still capped at the limit.
        • You need more than new tech, you need new physics to break that upper limit

          This is not a technological limitation, it's a physical limitation.
          The Mentalist
        • Actually...

          I read somewhere that scientists actually found a crystaline substance that light actually passed through at "faster than the speed of light" speed. One idea I can think of for bypassing this supposed limitation of computer speed would be a computer that went back in time to do its computations and spit the results out a fraction of a second after the process was initiated. And yes I realize time travel is not yet possible :D
          • Actually...

            they didn't. They used a quantum effect to propogate a fluctuation in a laser beam "faster than light." That this was theoretically possible has been known for some time. To use such an effect for any type of FTL transmission remains debatable, at most.
          • Apparently that effect is of no practical use...

            as you cannot use it to transport information at a speed faster than light.

            The speed of light is still the upper limit for information to travel.
            The Mentalist
        • Wrong.

          There is simply no reason to believe it will [b]ever[/b] be possible to break a fundamental limit imposed by the universe.

          Dream all you want of [i]faster-than-light[/i] or of [i]time travel[/i]. That's all it is: Dreams. The best evidence is not that we can't do such things, it's that such things cannot be done.
          • Do be careful about absolute certainty...

            ...after all short-sighted people are always saying "it can't be done, the laws of physics forbid it!"


            New laws of physics pop up like daisies every frew decades. Never is a *long* time...
          • And yet. . .

            . . . you are predicting what will be possible in the future based on unspecified new laws yet to be discovered.
          • But then again...

            I am basing my caution on past evidence.

            Everything someone says "it's impossible" or "it is certain that..." motivated humans either find a away around the problem or find a new approach that renders the problem meaningless.

            Given the accelerating pace of human discovery the next 75 years are likely to uncover far more scientific discoveries than the last 75 did. How much have we learned since 1934?

            Turing's paper on the Turing Machine wasn't published until *1937* after all...

            Besides, just like you I doubt I'll be around to be called on my prediction 75 years from now... :)
          • Our current understanding

            is that the speed of light is the ever present speed limit of the universe, to even approach it would do severe things to the very atoms of the human body. While I have no doubt of its role as a universal constant and a fundamental pillar of the universe, we have seen advancements in potential ways to travel faster than light without actually going faster than light, and some experimental mathmatics and physics are coming up with other quirks of the universe that may render this speed limit occassionally breakable. As for the computer speed limit, maybe we may need to sit at that speed for a while but we may someday learn to bypass it.
      • Software can help more yet but not Moore's law

        Software is now such a bloated inefficient thing it is constantly being criticised. MS is one of the least efficient software writers with so much bloated inefficiency. As hardware has accelerated operating systems and programs have slowed the technology down so that now PCs seem slower than a 486 did running DOS apps. I knew accountants who hung on to their 386 PCs until forced to change by Tax departments. Their PCs worked faster than the fastest Pentiums (at that time). So once we reach hardware limits we might just find the competition becomes focused on making software less bloated and inefficient. OSs run so much bloat, glitz, cross checks and record keeping they spend little time actually executing apps. Clean all this up and current PCs would SCREAM along like the old DOS systems did. The Cloud, if it becomes common use will slow us all down even more due to transfer times of data, programs etc so hardware speed won't even count.

        My point is that most of our current slowness is not the hardware. Current technology is capable of truly amazing speed already. Most users don't even need what we have now. Only research seems to need the speed and maybe some other unique uses. OK gamers want everything they can get but they are a unique breed and need to get a real life anyway.

        So all this focus on the ultimate speed is not warrented for probably 95% of users. Network and Internet speeds could be jacked up for sure.
        • It is hard ...

          "It is hard to make big things small, slow things fast, and complex things simple. Thus, there is no substitute for good design."

          One of my software teachers over 20 years ago said this. It seems no one teaches it anymore.
  • Still, the laws of physics are what they are.

    Still, the laws of physics are what they are. What happens when we get down to an atom? Then there's just a few electrons and protons and neutrons, but I find it doubtful that we'll get much smaller. Smaller than an electron?

    We may reach a point where the big discoveries are about fitting more chips into a 3D space. Like our brain - it's a large mass of computational power, not a wafer thin 2D square. But even then, there's only so much mass on this Earth, as big as it is.

    I think we'll reach limits. It's only a matter of when, not if.
    • Well Said

      Very well put. Thinking otherwise doesn't make scientific sense.
    • Size isn't relevant measurement for subatomic particles. [nt]

    • As they say.... laws are there to be broken

      And I am sure that they will find someway around
      the laws of physics. Heck, nothing is a law
      without an exception, they are even finding
      problems with some of Einstein's theories that
      were taken as laws up until very recently.
  • Yeah the same goes for the speed of light

    If you choose better fuel you may break it right? --end sarcasm

    Just read
    The Mentalist
    • Nope, but if you take a different approach

      then the speed of light no longer matters. Warp Drive, while once just science fiction, is actually theoretically possible. So your reference to the speed of light being an absolute limit is misleading even if it is correct. However if you move space around an object then you can exceed the speed of light without actually moving.

      There is nothing to say that our understanding of physics over the next 75 years won't completely change these principles. What if you had a processor that ran in multiple dimensions and/or universes? There are likely other universes where the laws of physics are completely different than ours and computing could be carried out on one of these other nearby membranes. There is just no telling.

      Nothing in science is absolute. While I think according to our current understanding of the universe and physics in general lends credance to what the scientests were discussing I also think it's silly to say "there is no possible way we will ever go faster than X".