IBM: Reinventing the Wheel at the Speed of Light

Moore's Law should stay relevant for a very long time. Every time it looks like we're hitting the limits of what the physics of semiconductor operation would allow, somebody pushes the boundary even further away.

IBM (NYSE: IBM  ) has done this a few times, with discoveries like strained silicon and silicon-on-insulator chip manufacturing technologies. Advances like these have helped Intel (Nasdaq: INTC  ) , Advanced Micro Devices (NYSE: AMD  ) , Texas Instruments (NYSE: TXN  ) , and others keep Moore's maxim of ever-increasing processor performance on track for more than four decades. Big Blue is at it again with research on how to make processors that use light instead of electricity.

The latest in a string of photonic computing breakthroughs, IBM's nanophotonic avalanche photodetector (say that one 10 times fast!) could be produced with standard tools of semiconductor manufacturing, using common materials like silicon and germanium. It aims to replace copper wires between computer chips with circuits that use pulses of light. No more than "a few tens of atoms" in size, the detectors could decode signals with so little power that a 1.5-volt battery might one day power a fully functioning computer. Oh, and IBM created the requisite ultra-fast light diodes way back in 2007. Try to keep up.

IBM Research spokesman T.C. Chen imagines that "the prospect of building power-efficient computer systems with performance at the Exaflop level might not be a very distant future." To put that vision into perspective, today's fastest supercomputer (a 37-000-processor monster built by Cray (Nasdaq: CRAY  ) for the Oak Ridge National Laboratory) achieves peak performance of 2.3 Petaflops, or just 0.23% of an Exaflop.

So if you've been holding off buying semiconductor stocks lately because they couldn't possibly get any better, well, you don't have that excuse anymore. As long as there's an Apple (Nasdaq: AAPL  ) building consumer-friendly electronic gadgets that increasingly demand more computing muscle with less power draw, a company like Pfizer (NYSE: PFE  ) running incredibly complex computer simulations of chemical and biological processes, or a government somewhere doing demographic analysis that makes you reach for the tinfoil hat, you can bet that there will be a market for faster, more efficient microchips.

How far will Moore's Law bring us before hitting the last immovable wall of physics? Pose your theories in the comments below. Me, I'm betting on double the performance every 24 months for at least 20 more years.

Fool contributor Anders Bylund owns shares in AMD, but he holds no other position in any of the companies discussed here. Intel and Pfizer are Motley Fool Inside Value choices. Apple is a Motley Fool Stock Advisor recommendation. Motley Fool Options has recommended a buy calls position on Intel. The Fool owns shares of Intel. Try any of our Foolish newsletters today, free for 30 days. You can check out Anders' holdings and a concise bio if you like, and The Motley Fool is investors writing for investors.


Read/Post Comments (4) | Recommend This Article (8)

Comments from our Foolish Readers

Help us keep this a respectfully Foolish area! This is a place for our readers to discuss, debate, and learn more about the Foolish investing topic you read about above. Help us keep it clean and safe. If you believe a comment is abusive or otherwise violates our Fool's Rules, please report it via the Report this Comment Report this Comment icon found on every comment.

  • Report this Comment On March 05, 2010, at 2:57 PM, gfbjohn wrote:

    Photonic computing? Wasn't that what was the big breakthrough that led to SkyNet being built in T1? :)

  • Report this Comment On March 05, 2010, at 3:00 PM, gfbjohn wrote:

    Hmmmmm, lessee... wasn't photonic computing the breakthrough technology that led to the development of SkyNet in T1?

  • Report this Comment On March 06, 2010, at 1:03 AM, jd22284 wrote:

    I worked on the wafers that were to be used as channels for them, its mind blowing the possibilities this opens in throughput... Mat. Sci. is the next IT

  • Report this Comment On January 22, 2011, at 5:56 PM, DDHv wrote:

    Didn't Moore's Law used to be a double every 18 months? Anyway, good going, Big Blue!

Add your comment.

DocumentId: 1128863, ~/Articles/ArticleHandler.aspx, 4/20/2014 2:51:00 AM

Report This Comment

Use this area to report a comment that you believe is in violation of the community guidelines. Our team will review the entry and take any appropriate action.

Sending report...


Advertisement