Tiny, tiny lasers
So where are the frikkin tiny, tink SHARKS!
Enquiring minds want to know!
Colin
<<< Icon 'cos I've had several :-)
The University of Texas at Austin has unveiled a laser built from a 28 nanometer rod and a 5 nm silicon dioxide layer. Despite the diminutive dimensions, the laser offers a lengthy stride It is, however, offering a good step forward in using optical communications between and within electronics. One of the challenges for a …
Thanks for the catch. However, the band gap of InGaN isn't that much different than InGaAs, so my question stands. A laser that size supporting a standing wave in the gain medium would either have to have one hell of a refractive index - like, 100 or so (way larger than any material of which I am familiar) - or would have to be operating in the extreme UV to soft X-ray part of the spectrum. And if these guys had made an EUV or X-ray laser at those scales they'd likely be booking a flight to Stockholm pretty soon.
an optical transistor eh? I believe you might be thinking of a reverse biased diode?
If the laser and the wave guides can be made small enough, there could certainly be a place for this in replacing global interconnects which are probably the biggest limiter to single core complexity/single thread perf. This wall was hit with P4 when they had to start pipelining wires. . .
So we have "gallium nitride nanoscale rod (only!) partly filled with indium gallium nitride". ???
And then we are told that laser run cooler than a piece of wire with a transistor at each end.
I am sure that this is important and exciting, but I am also sure the author didn't understand any of it.
I've always been suspicious about the notion that a micro laser for on chip data transfer would work out at lower power than straight electrons moving down a conductor, even a bad conductor.
I think sometime physicists forget to just *amazed* about the mismatch between the scales of the items they deal with.
Electron transitions generate visible light is a common observation people don't think twice about.
But an electron is c0.01nm and the wave length emitted is c500nm
IE 50 000 bigger. So far the best laymen's explanation seems to be that's just how quantum physics works.
It's impressive (and uses virtual particles to boot) but viable?