Intel Corp (INTC)

[SemiconEng]

The microelectronics lifespan might come to an end much sooner than some expect. All-optical chips (up to 1 MM higher speed then their microelectronics counterparts, and low noise) will replace them. The electrical wires inside the chip will be replaced by waveguides (the electrons will be replaced by photons).
The first elements (optical switches, and optical amplifiers, and so on) of this future trend are already here. See a sample below.
Mike


Maybe, and optical is very interesting technology IMO, but from what I've seen so far, in order to make optical chips mainstream, these guys are leaving out 2 very important concerns, that may be significant roadblocks.....

Manufacturablity: From what I hear, intel has been doing quite a bit of work in optical, but what I've started to realize recently, is getting new ideas out of the development lab, and into manufacturing, is a much larger issue then I use to think it was. Taking Low volume R&D of 1, 2, or even 5 wafer lots, with a handful of funtional die, and turning it into the yield numbers required for High Volume Manufacturing is MUCH more difficult then I was aware of..... 6 months ago

Cost: I'm also beginning to realize, although I guess it should have been intuitive, but in an R&D environment, nobody pretty much cares what it costs to get the technology to work. Nobody really even talks about the "Per Die Cost" in the meetings, which is very different then..... my previous experience. Sure optical works, but what's it going to cost? My guess is, that people might not want to hold their breath, expecting optical to even come close to parity costwise with current technology anytime soon, because holding their breath waiting for that, might cause suffocation

Don't forget, in the beginning, everyone was saying Gallium Arsenide was going to be the death of Silicon