It was back in the business section of the paper a few days ago. Some scientists at Intel had succeeded in creating tiny lasers on a computer chip. Then the article went on to speculate that this could perhaps lead to much faster communication between chips in a computer, and perhaps to solving the problem of economically bringing optical transmissions into the home.
Yes it will, but it will also lead to much, much more than that. This is a very, very big deal, akin to the moment 50 years ago when the first transistor was put on a chip. It will lead eventually to light computers, or in different words, computers of light.
Until now scientists have only been able to create "custom" lasers that were much larger, hundreds of times more expensive and/or made out of exotic materials.
The real breakthrough here is that tiny lasers can now be incorporated into a chip of silicon using fabrication techniques that are well within the range of current chip fabrication plants. This means that eventually there will not be just tens or hundreds of lasers on a chip, but millions or billions. And that will lead to the light chip, where the lasers are not simply working with transistors but have become the very transistors themselves.
Since lasers of light can be modulated on or off just like electrons traveling in silicon, these light transistors will be able to do anything that today's transistors can do, only at speeds of computation that are orders of magnitude greater that what is possible today.
It will also mean that computer memories will sooner or later vastly increase in density and storage capacity, since gigabytes of data can be loaded onto a single instant of a light wave. Circulating light waves within a chip could hold unimaginable amounts of data.
The other components necessary for a "light computer on a chip"—light diodes and light resistors—are already do-able. This means that though there are many technological challenges ahead before the light chip can become a reality, there are no inherently insuperable obstacles in that journey.
I'm not a scientist, just a layman. And no, I don't work for Intel. I don't even know anyone there. But mark my words, the recent breakthrough is equivalent to William Shockley's first silicon transistor chip in 1953, and will lead eventually to a technological world as distant from ours as ours is to 1953.
—jim sloman, 9.20.06
|
