Intel Aims For Light Speed
The New York Times' Bits blog has an interesting article about a new technology that Intel will soon be rolling out. It seeks to deliver data more quickly than possible on metal-based cables: at light speed using optical technology. If you've studied any physics, then you know nothing travels faster than light. For years I've been hearing that computers using optics, instead of wires, would ultimately be necessary to increase speed past what's currently possible. Intel is making progress in getting there.
The post explains:
The technology, which Intel calls Light Peak, will be enabled via a small chip and separate optical module.
There are a lot of reasons for going optical, the most obvious being speed: data can be delivered faster on optical cable than on current metal-based cables. Light Peak can carry data at 10 gigabits per second in both directions simultaneously, and Intel expects it will reach 100 gigabits per second in the next decade. By comparison, the ubiquitous Universal Serial Bus (U.S.B.) connectors on virtually all PCs today deliver data at a maximum speed of about 480 megabits per second -- or about 1/20th the initial speed of Light Peak.
How fast is that initial speed? It would transfer an entire Blu-ray movie in less than 30 seconds.
Through reading the piece, it's pretty clear that this is a significant opportunity for Intel. There are wide applications of optical technology in diverse fields of technology, computing and electronics. But everyone must be on board, according to Roger Kay, president of market research firm Endpoint Technologies, who Bits quotes:
"You have the TVs on one side, the PCs and cameras on the other. These industries are trying to make sure their little ecosystems stay stable. Theoretically, Light Peak makes perfect sense. But I do think there's a lot of politics that hold standards in place," Mr. Kay said.
The next generation of USB, 3.0., might provide enough performance improvements to satisfy most hardware manufacturers. And even in the best case, a transition to Light Peak will take years, Intel said.
I hope the transition doesn't get held up. A future where light is used, instead of metal cables, is inevitable. We might as well get there sooner than later. U.S. firms developing this technology first could also provide a nice global advantage in a sector where the U.S. desperately needs to remain relevant.