Sign in with
Sign up | Sign in

Engineers grow nanolasers on silicon

By - Source: UC Berkeley | B 15 comments

Engineers at the University of California at Berkeley said that they have found a way to grow nanolasers on a silicon surface.

The invention could lead to, if we believe the scientists, to a new class of much more efficient microprocessors.

Silicon has been known to be very deficient at generating light, which caused Berkeley engineers use another class of materials known as III-V semiconductors to create light-based components such as LEDs and lasers, even if there are substantial challenges: "Growing III-V semiconductor films on silicon is like forcing two in-congruent puzzle pieces together," said Roger Chen, a UC Berkeley graduate student. "It can be done, but the material gets damaged in the process."

The research group found a solution by growing nanopillars made of indium gallium arsenide, a III-V material, onto a silicon surface at the relatively cool temperature of 400 degrees Celsius."Working at nanoscale levels has enabled us to grow high quality III-V materials at low temperatures such that silicon electronics can retain their functionality," Chen said. He claims that technique could be applied to mass-manufacturing and  that it can generate near infrared laser light at room temperature.

"The hexagonal geometry dictated by the crystal structure of the nanopillars creates a new, efficient, light-trapping optical cavity," a press release stated. "Light circulates up and down the structure in a helical fashion and amplifies via this optical feedback mechanism."

There was no information whether and when this technology could go into production. the detailed results of the study are published in the February 6 issue of Nature Photonics.

Discuss
Display all 15 comments.
This thread is closed for comments
  • 2 Hide
    mister g , February 8, 2011 6:40 PM
    Saw this on PCWorld yesterday, wonder what'll come first this or graphene CPUs in mainstream PCs.
  • 9 Hide
    bto , February 8, 2011 6:43 PM
    intriguing! Now you really can have freaking sharks with freaking lasers attached to their heads!
  • 1 Hide
    jgalecio , February 8, 2011 7:01 PM
    I was thinking self powered nano lasers for nanoscopic surgery and even viral destruction!
  • 2 Hide
    dark_lord69 , February 8, 2011 8:09 PM
    jgalecioI was thinking self powered nano lasers for nanoscopic surgery and even viral destruction!

    Interesting...
    Although I think the goal of all this is to be able to allow a CPU to transmit data from one part of the CPU to another part using this technology which would work similar to fiber optics. I believe the ultimate goal would be to have the whole motherboard, CPU and RAM comunicating using connections that are similar to fiber optics to allow an unbelievable amount of bandwidth. But I could be completely wrong about all of that... :p 
  • 0 Hide
    fyasko , February 8, 2011 8:16 PM
    so this would be the next step in computer processing? or is it... "Marine diatoms, a unicellular algae, build their hard, patterned cell walls with microscopic lines of silica — a compound related to silicon, which is a key material for constructing computer chips and semiconductors."

    Read more: http://www.cbc.ca/technology/story/2008/01/21/algae-computers.html#ixzz1DPNQUj3v


    either way exciting stuff...
  • 1 Hide
    Grizely1 , February 8, 2011 8:19 PM
    making the cpu and mobo fiber optic wouldn't serve too much purpose. loss of data is the biggest reason for using fibre optic, specifically over distances (eliminating resistance in the transmitting medium)
  • 0 Hide
    Grizely1 , February 8, 2011 8:21 PM
    maybe in a few decades when cpus are bottlenecked by the speed at which signals transmit through copper mediums...
  • 2 Hide
    tpi2007 , February 8, 2011 9:29 PM
    "at the relatively cool temperature of 400 degrees Celsius."


    400 is not a relatively cool temperature in my book. Should it be 40 ?
  • 0 Hide
    Anonymous , February 8, 2011 9:38 PM
    Relative to a typical "hot" laser, 400C is cool.
  • 0 Hide
    bayouboy , February 8, 2011 9:42 PM
    Grizely1making the cpu and mobo fiber optic wouldn't serve too much purpose. loss of data is the biggest reason for using fibre optic, specifically over distances (eliminating resistance in the transmitting medium)


    Uhhh, physics says otherwise man. Light is much faster than the propagation of a magnetic field in an electrical circuit which is much faster than electron flow. So yes, using light would yield a significantly faster CPU and far better bandwidth all around.

    If silicon was as fast as light, it would have to be superconducting.

    Are you saying that your CPU has superconducting material?!?
  • 0 Hide
    fans 6 , February 8, 2011 9:59 PM
    Engineers t Berkley? LOL. Probably getting paid 1/4th of an Intel Engineer.
  • 0 Hide
    dalauder , February 8, 2011 11:19 PM
    bayouboyUhhh, physics says otherwise man. Light is much faster than the propagation of a magnetic field in an electrical circuit which is much faster than electron flow. So yes, using light would yield a significantly faster CPU and far better bandwidth all around.If silicon was as fast as light, it would have to be superconducting.Are you saying that your CPU has superconducting material?!?

    Dude, I think he's just saying that heat(and other things) in the CPU is our bottleneck, not the fact that the mobo uses copper.
  • 0 Hide
    JPForums , February 9, 2011 12:06 PM
    bayouboyUhhh, physics says otherwise man. Light is much faster than the propagation of a magnetic field in an electrical circuit which is much faster than electron flow.

    Actually, it is the propagation speed of an electromagnetic wave that affects the transferring of information.
    Electromagnetic waves propagate at the speed of light in a vacuum, 95% the speed of light through unshielded copper, and 66% the speed of light through coaxial (shielded copper).

    Drift velocity which is caused by an electric field is far to slow to be of use in data transmission.

    While there is a potential speed up of up to 50% en route, the generation and detection of light has to be as fast as that of electricity in our current circuits to be of use for small runs.
    This is especially true in electric/laser hybrid circuits where there is an additional latency in transforming electricity to light and vice versa.
    Longer runs stand to gain more because more time is spent en route.
    I would expect to see this technology on consumer motherboards before seeing it on consumer processors.
    Of course, other factors, I.E. cost and ease of implementation, could dictate otherwise.
  • 0 Hide
    christop , February 9, 2011 2:24 PM
    Cool..
  • 0 Hide
    ubercake , February 11, 2011 10:37 AM
    Some seriously insane achievements! Wow. Really cool.