Forgot your password?
typodupeerror

Intel Announces Lasers On a Chip 244

Posted by kdawson
from the photonics-awakes dept.
wonkavader writes, "The New York Times reports that 'Researchers plan to announce on Monday that they have created a silicon-based chip that can produce laser beams. The advance will make it possible to use laser light rather than wires to send data between chips, removing the most significant bottleneck in computer design.' The work is from Intel and the University of California, Santa Barbara. This suggests breakthroughs in both computing performance and networking." From the article: "The breakthrough was achieved by bonding a layer of light-emitting indium phosphide onto the surface of a standard silicon chip etched with special channels that act as light-wave guides. The resulting sandwich has the potential to create on a computer chip hundreds and possibly thousands of tiny, bright lasers that can be switched on and off billions of times a second." Further details in the Intel press release.
This discussion has been archived. No new comments can be posted.

Intel Announces Lasers On a Chip

Comments Filter:
  • by base3 (539820) on Monday September 18, 2006 @01:17PM (#16131328)
    . . . to be announced shortly.
  • Tron (Score:5, Funny)

    by pythiane (1003082) on Monday September 18, 2006 @01:18PM (#16131350) Homepage
    And Tron is yet another step closer to fact.
  • About time (Score:5, Interesting)

    by dorpus (636554) on Monday September 18, 2006 @01:20PM (#16131369)
    They've been trying to build optical computing chips since the 1980s. I did a summer internship in Japan in 1990, when they were making custom batches of exotic rare-earth crystals for fiber-optic relay stations.
    • by suv4x4 (956391)
      They've been trying to build optical computing chips since the 1980s. I did a summer internship in Japan in 1990, when they were making custom batches of exotic rare-earth crystals for fiber-optic relay stations.

      I was going to say the same. They keep trumpeting their chip laser technology on every keynote I've seen for years now.

      There's a flaw with using lasers for integral schemes: they go in a straight direction, wires can "steer" and form more complex patterns. Of course lasers can also cross each other
      • by KDR_11k (778916) on Monday September 18, 2006 @05:01PM (#16133596)
        There's a flaw with using lasers for integral schemes: they go in a straight direction, wires can "steer" and form more complex patterns. Of course lasers can also cross each other and wires can't.

        May I introduce you to a groundbreaking new technology called "glass fibres"?
  • by dmatos (232892) on Monday September 18, 2006 @01:21PM (#16131385)
    For blue LEDs used by case modders. Why bother when the chips are flashing all by themselves.
    • Re: (Score:3, Funny)

      by sl3xd (111641) *
      I can see it now:
      * Blinky new CPU: $1000.00
      * Transparent heatsink made of Aluminum oxynitride: $5,000.00
      * Being the 1337357 h4x0r in the whole basement: Priceless.
  • Switching (Score:3, Insightful)

    by Zebadias (861722) on Monday September 18, 2006 @01:21PM (#16131387)
    I think this will be of more use to optical switching - if you have the ability to switch and route on your fibernetwork without changing from optical to electrical and back again you can switch much faster and more efficiently.
    • Re: (Score:3, Insightful)

      by jimstapleton (999106)
      the laser is still being generated by the chip (and hence, I suspect, by elecrtical), so I don't think that works.
    • Re:Switching (Score:5, Informative)

      by Relic of the Future (118669) <dalesNO@SPAMdigitalfreaks.org> on Monday September 18, 2006 @02:31PM (#16132073)
      No, this isn't optical switching. Laser light still comes in, gets converted to electrons, calculations are performed, and then more laser light is generated and sent out.

      What this does is make it much simpiler (and CHEAPER) to make the laser light, to the point where it's worth while to have a fiberoptic connection between, say, your CPU and and your vRAM, or between your IDE controller and your RAM, rather than the terribly capacitive and inductive (and therefore SLOW) motherboard trace.

  • Go Intel! (Score:3, Funny)

    by Cybert4 (994278) * on Monday September 18, 2006 @01:21PM (#16131388)
    Great company. Real solid and with great integrity. I'm sure they'll put lasers to great use. Yes, x86 is horrible, but that too will pass.
    • Yes, x86 is horrible, but that too will pass
      That ship has already struck the iceberg. Intel misses [2004] Itanium sales mark by $26.6bn [theregister.co.uk] And you thought your sales projections sucked.
    • by peragrin (659227)
      Yes, x86 is horrible, but that too will pass.
       
      Not as long as MSFT is making the default OS. Heck MSFT can't even get away from the now ~25 year old tech of BIOS and floppy disks. Vista will finally kill the need floppy's but MSFT refuses to fully support EFI. You have to use a bios compatiblity module to boot vista.

      Yet Apple is on their second major system architecture change in 15 years. and Linux. Well linux runs on nearly anythng.

      • Re: (Score:2, Informative)

        by Cybert4 (994278) *
        Uh, Apple uses Intel. Heard? Also, you don't need floppies for XP. Except for some corner case--you can boot CD's or thumb drives just fine.
        • If you're doing anything interesting with your hard drives (RAID, JBOD, whatever), you need a floppy drive to install XP. Big pain in the ass.
          • by 2short (466733)

            I installed XP on a RAID on my home machine that doesn't have a floppy drive. I just booted from CD.

            Certainly the rack mount/blade servers the IT guys down the hall don't waste sppace for a floppy drive, and they certainly do all the "interesting" stuff with hard drives.
      • I don't think it's all Microsoft's fault, but PCs do need to much legacy nonsense. I mainly use Macs. I am building my own computer to run a MythTV server, and I find having half of my ports completely useless quite irritating. There is also the need for floppies to do some things. Parallel, serial, and COM ports need to go. Floppies need to go. PATA HDs need to go. Why can't hardware manufacturers be brave, and get rid of that junk?
    • by eclectro (227083)
      Great company. Real solid and with great integrity. I'm sure they'll put lasers to great use. Yes, x86 is horrible, but that too will pass.

      User criticized x86. Laser firing on user Cyber14 in 4...3...2...
  • by ZachPruckowski (918562) <zachary.pruckowski@gmail.com> on Monday September 18, 2006 @01:22PM (#16131400)
    Obviously this boosts bandwidth and cuts latency (like mad), but doesn't this kill the current FSB speed and multiplier method? I mean, your clock speed is FSB clock x multiplier, so what happens if you replace the FSB with a laser?
    • Re: (Score:3, Informative)

      by jimstapleton (999106)
      nothing, instead of EM pulses propigated by electrically conductive substances, it will be self propigating photons directed by optics.

      If I'm reading it right, most of the control could be handled by the same mechanisms, it's just that different signal senders and recievers will need to be used.

      And, I thought lasers didn't offer significantly lower latency, only better bandwidth?
      • by dgatwood (11270) on Monday September 18, 2006 @03:31PM (#16132730) Journal

        The speed of electrical propagation in copper (~200,000 km/sec) is about 2/3rds the speed of light in a vacuum (~299,792 km/sec). Think of it as having about 2/3rds the latency of copper and you'll be about right, assuming the light goes through open air.

        Now if you mean light through an optical cable, it's about as slow as a signal through copper, so there's no real gain.

        The real benefit here is short interconnects without any medium in-between. CPU vendors have done this within chips by putting edge contacts on cores so that they can tessellate the cores and have them connected together. With optical edge connects, the failure rate will be lower because the contacts won't corrode and don't have to be soldered.

    • Re: (Score:2, Informative)

      by LiquidCoooled (634315)
      This looks more for CPU interconnects than for actual CPU processing.
      The data still has to be transmitted and still has to get back.
    • Re: (Score:3, Insightful)

      by Sebastopol (189276)
      Pushing the FSB is much harder due to the platform. The physical interconnect is far noisier than on die routing, and the distribution of those signals to the memory and/or IO controller is very messy. That's why FSBs are so much slower (or if they are faster, or usually dedicated point-2-point busses).

      To reap the benefits of optics outside the package you'd need an optical socket and a radically new kind of mobo design.

      Give it 20 more years...
    • Then the FSB can run as fast as the chip; no need to slow the signal down to send it across that old slow electrical bus. You're not "losing" a multiplier, you're losing a DIVISION.
    • by raddan (519638)
      I am not an ECE, but I'm guessing that the components on the bus still need a clock to talk to each other.
  • A huge advance? (Score:4, Interesting)

    by Coppit (2441) on Monday September 18, 2006 @01:28PM (#16131467) Homepage
    From what I recall in physics class electrons travel at 2/3 c. So at best this means that memories and chips can be 50% further apart, or that clocks can go 50% faster. Or is there more to this?
    • by rbarreira (836272)
      Heat maybe? I'm not sure either...
    • Re: (Score:2, Informative)

      by P3NIS_CLEAVER (860022)
      When you are talking about electrons you start to have problems with resonance and interference between the connections. This is why memory is such a difficult problem, because manufaturer A has to create a memory module that plays nice on the generic memory bus designed by manufacturer B. If there is an optical buss from the CPU to the memory module, the memory manufacture has carte blanch to design a module as fast as they want, because there is no more buss restrictions. They would only have to solve th
    • Re:A huge advance? (Score:5, Informative)

      by MightyYar (622222) on Monday September 18, 2006 @01:45PM (#16131662)
      I'm in the industry, but this isn't my specialty. From what I remember, the speed of the electrons isn't why this is important. There are electromagnetic effects that limit the speed of communications... things like crosstalk. The little balls, wires, or deposited metal that they currently use to make the interconnections are like tiny little antennas. The interconnections are also a pain in general, no matter what technology is used, because of things like thermal mismatches and encapsulation problems. From a packaging standpoint, this would solve many problems, and probably create even more - alignment, anyone?
      • Re: (Score:3, Informative)

        by Steve525 (236741)
        People are concerned about bandwidth, not speed. I.e. how much data can you put down a wire (and how big is the wire). Or, at least bandwidth is the only thing they can hope to improve, electrical signals already travel pretty close to the speed of light. Part of what limits electrically lines is RC limits - frequencies beyond resistance * capacitance can't travel. Any line is going to have finite capacitance and resistance. In addition, there may be dispersion and other effects causing high frequency
        • by MightyYar (622222)

          That sounds better than what I said :)

          I'm in the chip assembly equipment end, and the main problem that I think they will have is lining up the chip with the substrate so that the lasers point at the detectors. The additional challenge will be keeping them lined up, since the substrate usually has a different coefficient of thermal expansion (COE) than the chip. Even if the COE matches, the temperature may not. Presumably, they will have to wire up the power to these chips, perhaps using a wire bonder, an

      • Power (Score:3, Informative)

        by stevesliva (648202)
        One huge advantage could be an orders-of-magnitude reduction in the current necessary to drive signals off-chip. (It's not mentioned in the article whether these drivers have a power advantage) Off-chip drivers are a significant source of current drain in a chip, and if this technique eliminates the necessity to wiggle the off-chip capacitive loads at high frequencies, then you'll see much lower power. And if each pin on the output bus is drawing less power, you may see larger bus sizes and more bandwidth
    • by Moby Cock (771358)
      Many advantages. Reduced ohmic losses will make chips run cooler. Or they can run faster for the same TDP. Also, the switching speeds tend to be faster for optical transmissions compared to silicon transistors. Fibre optic networks can carry higher bit rates than twisted pairs or coax.
    • by ceoyoyo (59147)
      Actually, electrons in a wire physically travel at about a walking pace. You can transmit a signal using them at about 2/3c, which is what we're interested in of course.
    • Bosons vs. Fermions (Score:5, Informative)

      by monopole (44023) on Monday September 18, 2006 @03:05PM (#16132457)
      Electronic signals travel pretty damn close to c. The problem is that electrons are fermions and as a result are antisocial by the Pauli exclusion principle no more than 2 in each location. Charge makes this even worse. On the other hand photons are boson and they like to hang out in the same location. As a result electrons are handy when you want bits to interact (logic gates, memory) while photons are handy when you want bits to pass through each other (communications etc.). The advantage of using photons is that you can make connections without EMI or other cross talk problems. In addition there is some very nifty quantum computing you can do with such systems (the topic of my dissertation).
    • Re: (Score:3, Informative)

      by wass (72082)
      As others have noted, the drift velocity of electrons is fairly slow, this is the average speed a single electron will migrate along a device in the presence of an electric field (ie, in an applied voltage across the device).

      However, a changing voltage signal will propagate at speeds of order c (smaller than c, of course). The 'wires' ror traces unning on the microprocessor are basically transmission lines, so you're really transmitting electromagnetic signals. This is just like standard textbook transm

    • Re: (Score:3, Informative)

      by Ruie (30480)

      From what I recall in physics class electrons travel at 2/3 c. So at best this means that memories and chips can be 50% further apart, or that clocks can go 50% faster. Or is there more to this?

      The simplest way to explain this is to note that a wire is an inductor - and at high frequencies this matters. What is more, a 1Ghz digital signal needs bandwidth much larger than 1Ghz - or the edges of ones and zeros get distorted too much. If CPUs used analog signals inside to transmit information between chips (

  • Safe? (Score:5, Funny)

    by pafmax (462211) on Monday September 18, 2006 @01:29PM (#16131475) Homepage
    The future of IM:
    - Hey look at what I'm sending you!
    - ARGH! MY EYES!!!

    Seriously, are these lasers safe?
    • by owlstead (636356)
      Depends, it would probably be strong enough to roast an amoeba. So if you're one, I would be scared, very scared.
    • Re:Safe? (Score:5, Funny)

      by reverseengineer (580922) on Monday September 18, 2006 @02:03PM (#16131817)
      The future of IM:
      - Hey look at what I'm sending you!
      - ARGH! MY EYES!!!

      That's pretty much what IM is like now.

    • The *future* of IM? With goatse and tubgirl, the future is NOW!
    • by Moby Cock (771358)
      I assume you mean eye safe? If so, then the answer is yes. These lasers are on the chip! And are designed to transmit between two points on the chip. Just as there is almost no danger of electrocution from a contemporary chip, these will post no danger of blinding anyone.
    • by monopole (44023)
      If they put out enough power to be a threat their heat dissipation would destroy the chip. About as dangerous as a spread out red laser pointer
    • Re: (Score:3, Funny)

      by vmxeo (173325)
      WARNING: Do not look at laser chip with remaining eye
  • New Techniques... (Score:3, Interesting)

    by skogs (628589) on Monday September 18, 2006 @01:31PM (#16131503) Journal
    This makes me wonder about the future new techniques this could be used for. Never mind the obvious inter-chip communication...how about visual systems?

    Could this, with another 10 years of evolution and the advancement of color coordination and multi-colored laser chips, provide incredibly high contrast and accurate projections? This is like DLP projectors on steroids. They don't simply reflect light one pixel at at time, they actually create the laser one pixel at a time.

    I also was wondering what the 3D applications would be like. Perhaps an R2D2 unit fitted with one of these would have a much sharper and sexier image of the princess asking for OB1's help.

    Also, how about a laser weapon targeting system that can lase 100 targets at once for all the bomblets?

    Great things are going on in my mind.
    • Re: (Score:3, Funny)

      by Elwood P Dowd (16933)
      Great things are going on in my mind.
      It's like a laser.
    • I also was wondering what the 3D applications would be like. Perhaps an R2D2 unit fitted with one of these would have a much sharper and sexier image of the princess asking for OB1's help.

      Also, how about a laser weapon targeting system that can lase 100 targets at once for all the bomblets?

      Great things are going on in my mind.


      Okay dude, I was with you when you talked about a sexier picture of Leia, but the moment you talked about weapons, I could no longer support the idea that your mind was great.
  • Put a lot of people who know a bit about computer science (linux, PHP, etc. ) and have them comment on a hard science. They don't even know enough hard physics and math to even rate their own skills. All they can do is joke about it. Enough with the sharks.
  • To me, this sounds like an extremely high-res display device. I wonder how suitable this stuff is for displays anyway? Will the manufacturing process be easier/faster/cheaper than other technologies in use today? Will it open doors for the creation of live 3D holographic displays?
  • I just saw this. (Score:5, Informative)

    by Steve525 (236741) on Monday September 18, 2006 @01:53PM (#16131732)
    I was at a conference last weel (http://www.ieee.org/organizations/society/leos/LE OSCONF/GFP2006/index.html [ieee.org]) were this was presented by John Bowers. As they explain briefly in the article, they are bonding InP to Silicon wafers. The silicon provides the waveguiding, and enough of the mode is in the InP to give them gain. They achieved an optically pumped laser, and were still working on an electrically pumped one. I wonder if this announcement will mean that they achieved electrically pumped lasing.

    It's good work, but I'm not sure if the bonding process will ever be suitable for monolithography integrated CMOS and photonics. I was more impressed by the work done in Huffaker's lab (http://www.chtm.unm.edu/huffaker/index.html [unm.edu]) where they are working on growing III-V materials directly on silicon. However, the work by Bowers is more mature and will lead to devices sooner.
  • 50% of the visible comments today are regurgitated fad material... jeez...
  • by baggins2001 (697667) on Monday September 18, 2006 @01:58PM (#16131768)
    I be it will take at least 5 to 10 years to see this on a standard desktop/server system.
    My biggest concern is reliability. How many people are running SANS with redundant Fiber optic connections. Why? because the lasers fail. Could you imagine if you had a motherboard built with multiple lasers for on board communication. Yeah it would be fast, right up until the time one of those lasers failed.
    InP lasers on silicon is new technology and is quite a ways from being producible in a mass market chip. Manufacturers have enough trouble getting gates, isolation, contacts for silicon devices reproduced. Now tell them to create a step where they put a laser in there and I bet it will take them 2-3 years design and 3 years to get a manufacturing process. (Can anyone say copper level metal?).
    Hopefully this isn't something that completely patentable, because this is where the consumers would benefit from competition.
    From a manufacturing perspective, I would rather be stuck trying to implement TaO gates.
  • by MetaDFF (944558) on Monday September 18, 2006 @02:20PM (#16131967)
    By using optical links, this breakthough will solve some of the problems we have today with sending data at high speed across chip to chip busses. The major problem today with sending data at high rates between chips is the losses incurred by travelling across the FR-4 PCB. As the data rates go up, the greater the losses incurred, the more difficult it is to recover the data being sent. Optical interconnects have significantly less losses at high data rates, thus making them a suitable technology for chip to chip communications in the future. This is a breakthrough because now we can integrate exotic optical materials with low cost silicon using standard chip-making equipment. This was something that could not be done in the past.
  • Figure out how to mount those chips onto the heads of sharks...

  • I'd go for lasers on a chip. It's got to be better than Snakes on a Plane.
  • Power anyone? (Score:2, Interesting)

    by lixee (863589)
    I just had a course on advanced VLSI design, where the Professor relies on [Kibar, VanBlerkom, Fan, Esener, J Lightwave Techn., vol. 17, p. 546, 1999] to approximate a couple of Watts for optical interconnects. This is clearly not acceptable.
    I'm interrested in how they manage to keep the power consumption reasonable. Till then, I call hype!

  • First, bubble memory, then optical processors, then...

    ...flying cars!!!

    Yay!!

    MjM

  • by smellsofbikes (890263) on Monday September 18, 2006 @03:16PM (#16132560) Journal
    1. Why lasers? Why not just light? At the distances they're talking, does coherence and phase matter? Incoherent light is just as fast, and if you're shooting it into waveguides and it's coming out the other end, as long as you're not multiplexing data on a given waveguide what advantage does this give? (I honestly don't know: maybe there's a great reason.)

    2. They're still bonding indium phosphide onto an existing chip. When they can use photolithography to build a billion lasers on the chip itself, rather than having to glue separate lasers onto a chip, that'll be really impressive. That's why so much effort is being focussed (pardon me) on developing silicon lasers [brown.edu] rather than exotics attached to silicon.
    • "as long as you're not multiplexing data on a given waveguide what advantage does this give?"

      The ability to multiplex data on any given waveguide (ie: boost bandwidth per lead)

      • It makes sense to multiplex if you have lots of leads connected -- a bus, in other words. Does it make sense from silicon-to-lead?

        It seems to me that the whole point of optic computing is to replace the whole chip with waveguides and do all the computing with light: no crosstalk, no capacitance. What Intel has done is keep the chip, but swap out the bondwires with optical ones. That's pretty cool but it's not obvious to me that it beats current interconnects. Yeah, it's neat that they can multiplex, but
    • Re: (Score:2, Funny)

      by Slicebo (221580)
      "At the distances they're talking, does coherence . . . matter?"

      You're posting to **Slashdot**. You gotta know the answer to this one.
  • >> Researchers plan to announce on Monday that they have created a silicon-based chip that can produce laser beams.

    Isn't this just what laser LED's are? They've been around for years now.
  • I thought the headline read "Intel Announces Lawyers On a Chip", and I was really scared for a moment. God forbid Intel unleash billions of transistor-powered attorneys on the world!
  • So, will the motherboards on our computers need to be sealed in an air-tight environment to keep out dust? I imagine the beams need to be fairly small in diameter to make this practical.

    Also, could the removal of electronic pathways in place of lasers result in a much cooler running system, since there won't be heat generated by electronic resistance?
  • Intel announces "Motherfuckin Lasers on a Motherfuckin chip"

    sri
  • Vaporware. It doesn't exist yet. It won't be in production for 10 years. It is a non-product and of no consequence at the moment.
  • About 20 years ago I remember reading an IEEE publication that talked about doing this very thing, and i dont think it was a prediction as it had pictures.

  • Has someone bothered to calculate how long it takes to go from press release to Best Buy for new technologies in different areas?

Don't hit the keys so hard, it hurts.

Working...