Moore's Law, based on an accurate prediction in a 1965 article written by Intel
UCLA and Intel have independently made several announcements claiming big advances in building a new laser and another communication device called a modulator. This may initially seem like a big yawner; companies like Coherent
The silicon story
Silicon is the world's most popular semiconductor material, but its use has been limited to electronic applications -- those that depend on the flow of electricity. Despite all its success in electronics, silicon has been a big flop in optical applications like communications. These applications have been "owned" by more exotic (and expensive) materials such as gallium arsenide. Costs of computing and communications would be reduced significantly if silicon could be used instead; it's a cheaper material with an extensive existing manufacturing base that could be used to create optical components.
Silicon hasn't broken into optical communications because it's a poor light-emitter -- and building a laser out of a material that only grudgingly emits light is a tough task. But that's only part of the problem. In optical communications, the laser beam conveys data by varying its brightness in some predictable way. If the beam doesn't vary, it can't carry information. Changing the brightness requires a device called a modulator; the challenge here is making a modulator fast enough to pack a lot of data onto the beam in a short time. In this respect, silicon has also come up short.
In early 2004, Intel reported that it had developed a fast modulator using silicon. Like all modulators, Intel's worked somewhat like the mechanical shutter on your camera -- if it were able to open and close a billion times per second. This sounds fast, but it still wasn't fast enough. Happily for Intel, earlier this year the company announced a modulator that was 10 times faster.
In October 2004, UCLA researchers reported the development of the first silicon laser. They found a way, dubbed the Raman effect, to coax laser light out of silicon using a method that is employed in optical fibers to amplify light. This first laser was pulsed, meaning that the light output was chopped up into little bits, kind of like repeatedly flipping a light switch on and off. The UCLA laser operated only in pulsed mode because it caused a buildup of electrons that absorbed the laser light. Trying to sustain a laser beam when the light is simultaneously being absorbed is like paddling a canoe upstream -- the harder you paddle, the stronger the opposing current becomes. Fortunately, the electrons go away by themselves if you wait a bit -- hence the reason for pulsed operation. Nevertheless, continuous lasers are preferred in communications, and modulators are used to chop up the beam. Even though UCLA's silicon laser was a big breakthrough, more work was needed.
Not surprisingly, silicon's potential for use in lasers had not escaped the attention of Intel. In February 2005, Intel researchers reported the first continuous laser made from silicon. Intel was able to prevent the buildup of electrons by using voltage to sweep them out of the way -- kind of like the wind sweeping leaves out of your driveway.
More technical challenges remain, but it now looks much more likely that silicon optical components, or silicon photonics, will eventually become a reality. Its eventual applications may even reach as far as your home computer.
You might have heard that Intel recently released a dual-core processor. These chips have two identical execution cores housed inside one processor. Currently, for one core to communicate with the other, data must pass through a copper wire linking the two cores. This works just fine, except that it's slow. What good does it do to perform calculations at lightning speed if the results of the calculations just dribble out like ketchup from a bottle? The need for high speed is why communications networks now use optical technology. The signals are converted to light and sent down an optical fiber because optical signals have much higher
frequencies than electrical signals, allowing them to carry more information. Similarly, in a dual-core processor, silicon photonics built into the chip could dramatically increase communication speed between the cores.
Even better news: Since Intel and Advanced Micro Devices
Silicon optical components may also find applications in networking. It's possible that manufacturing costs would fall so much that each of us could install a fiber-optic network in our home to deliver high-speed data, including high-definition video for all of our televisions.
It is certain that many more currently unimagined applications will be found for this technology. The era of silicon electronics has changed our lives in ways that were completely unpredictable 50 years ago. Perhaps silicon photonics will one day have a similar impact.
For more laser-sharp Foolishness:
- Is II-VI Still Beaming?
- Shedding a Little Light on Rofin-Sinar
- Candela's Lasers Offer Dim View
- Light up the conversation on our Intel message board.
High tech, biotech, nanotech -- if it's on the cutting edge, it's covered in Motley Fool Rule Breakers. David Gardner's newsletter service scours the world of next-generation technology to find stocks with explosive growth potential. Ready to start investing in companies that just might change the world? Sign up today for a free 30-day trial.
Dan Bloom enjoys writing about the stock market and technology, but he isn't exactly an early adopter. He even went for months without a cordless phone when the old one broke. He doesn't own shares of any stocks mentioned in this article. The Fool has a disclosure policy.