Forgot your password?
typodupeerror
Shark Hardware

Scientists Build Three Atom Thick LEDs 54

Posted by Unknown Lamer
from the smaller-better-faster dept.
minty3 tipped us to news that UW researchers have built the thinnest LEDs yet: a mere three atoms thick. Quoting El Reg: "Team leader Xiaodong Xu, a UW assistant professor in physics and materials science and engineering, and his graduate student Ross, have published the technique in the latest issue of Nature Nanotechnology. They report that the LEDs are small and powerful enough to be used in optical chips that use light instead of electricity to shuttle signals and data through a processor, or they could be stacked to make new thin and flexible displays."
This discussion has been archived. No new comments can be posted.

Scientists Build Three Atom Thick LEDs

Comments Filter:
  • by Rick in China (2934527) on Tuesday March 11, 2014 @12:17AM (#46452315)
    When I see it!
  • by hey! (33014) on Tuesday March 11, 2014 @12:23AM (#46452325) Homepage Journal

    Unfortunately, that means they have to be several kilometers in width...

  • Insulation... (Score:4, Insightful)

    by Miamicanes (730264) on Tuesday March 11, 2014 @12:45AM (#46452373)

    And how many atoms thick does the insulating layer between adjacent photosensitive or photoemitting structures need to be to prevent light emitted by one pair's LED from unduly influencing the state of an adjacent photodiode/phototransistor?

    What, exactly, is the benefit of building a chip whose internal connections are basically all optoisolators?

    • by Mashiki (184564)

      What, exactly, is the benefit of building a chip whose internal connections are basically all optoisolators?

      Besides omnidirectional communication without the need for direct pathways. Oh...not much.

      • Re:Insulation... (Score:4, Informative)

        by BitZtream (692029) on Tuesday March 11, 2014 @11:08AM (#46454667)

        We've had that for years already, and no, its not all that useful in computer chips. You need very specific directed communications almost exclusively except for the clock pulse.

        • by Mashiki (184564)

          We've had that for years already, and no, its not all that useful in computer chips. You need very specific directed communications almost exclusively except for the clock pulse.

          Sure, it's not useful in modern chips however. Following the theories on optical chips however, you don't need it to be directed when banks are set projected.

    • by Anonymous Coward

      What, exactly, is the benefit of building a chip whose internal connections are basically all optoisolators?

      Light travels faster than electricity, and with less interference. Since a critical limitation of modern chips is that it takes more than 1 clock tick for data to travel from one side of the chip to the other, this is of mindblowing importance.

      And how many atoms thick does the insulating layer between adjacent photosensitive or photoemitting structures need to be to prevent light emitted by one pair'

    • by Anonymous Coward

      Modern CPUs are starting to face a bandwidth and power limitation for their interconnects. One solution that is being investigated is using photonics, because light can carry much more data (several THz of bandwidth) vs electrical interconnects (10s of GHz). An on-chip light source would be an important component for this.

    • Re: (Score:3, Insightful)

      As other people have said, the advantage is speed.

      Of course nobody is expecting this tech to replace silicon based chips anytime soon. There's obviously a lot of R&D to be done and, let's face it, nothing may EVER come of it. That's just how science is. We don't know in advance what theories and tech will pan out.

      As for leakage between structures? I'm willing to bet we don't need perfect isolation. Just enough isolation that the interference is predictable. (Much like electrons in silicon...)

  • by CrankyFool (680025) on Tuesday March 11, 2014 @12:45AM (#46452377)

    I'm impressed that they didn't just build one one atom thick LED, but three of them. Was it to prove they could reproduce it?

  • by enos (627034) on Tuesday March 11, 2014 @01:35AM (#46452511)

    To be clear, only two authors are from the University of Washington. They have many collaborators, including from Univ. of Tennessee, Oak Ridge NL, Germany, Japan, and Hong Kong.

    Submitter: University of Wisconsin and University of Waterloo are also known as "UW". It's worth expanding on first use.

  • So the electrons from how many atoms cross the junction?
  • "Scientists Build Three Atom Thick LEDs"
    Why didn't they just build one ?
  • Does anyone else remember when IBM would be the one to do something like this?

  • by Anonymous Coward

    There are two other publications in the same issue of Nature Nanotech., both of which also report the creation of similar systems also utilizing monolayers of WSe2: this one [doi.org] from a group at the Vienna University of Technology and this one [doi.org] from a group at MIT.

  • So, are we back in the game?

    We can observe clear evidence that Moore’s Law is ending, because we can point to a pattern that precedes the end of exploiting any kind of resource. But there’s no reason to panic, because Moore’s Law limits only one kind of scaling, and we have already started another kind.

    • by qpqp (1969898)
      Oops, URL missing: http://herbsutter.com/welcome-... [herbsutter.com]
    • by Megol (3135005)
      Nope. Moore's law is an observation of transistor growth. Doing optical communications have nothing to do with it at least to a first approximation.

      That said optical communication does open up a lot of advantages including (potentially) reduced signal latency, denser and thus higher bandwidth chip I/O and other effects that will (if practically usable) increase performance of computers. But transistor scaling? It's dead.

      • by qpqp (1969898)
        So would it be more correct to say that we're back in the exponential game on single chips (using Moore's law only by inference), as opposed to horizontal scaling by adding cores?
  • Hmm, 3 atoms for the LED and a million atoms for each solder joint.
  • Is it a single LED that is three atoms thick, or are there 3 individual atom thick LEDs?

    If the latter, 1 atom thick is just an atom. So these guys are claiming to have invented an atom.

System checkpoint complete.

Working...