Seems like a banner day. First stem cells, now Intel is announcing in Nature they have found ways to build fiber optics into chips using fairly conventional materials and techniques. This should make it possible to bring down the cost of high-bandwidth telecommunications gear radically.
Paul S. Otellini, the president of Intel, said on Wednesday: “Think of it either bringing us a tenfold decrease in costs of existing communications, or 10 times the bandwidth for the same cost. It is yet another step in the path to convergence we have been discussing for the past few years.”
So what? Well, Moore’s Law shows faster, smaller, cheaper stuff makes a big difference in availability and impact. Already existing “cheap” broadband has been credited–or blamed, depending on point of view–for fueling the offshoring of work. Think of all the rather lame video, audio networking things we have today put on steroids.
“There is a big difference between 10 megabits and one gigabit,” Terabit’s Dr. Huang said, noting that the advance should ultimately make it possible to create computers that “will be everywhere that you are.”
Hmm. How come that doesn’t seem like an unalloyed good to me?