| brandon - Feb-11-2004 server time |
| 2 microns is not that big! |
| VictorEM83 - Feb-11-2004 server time |
| A quote from the artical “Actually, these devices are not that small. In the lab we have scaled to smaller dimensions that we aren't ready to discuss yet. This (1-GHz) device has a rib waveguide in the polysilicon area that is 2.5 microns wide and roughly 2.3 microns high,” said Paniccia. Sounds like the interface to the laser is tiny then the laser itself can be bigger outside of the chip |
| Red Squirrel - Feb-11-2004 server time |
| Yes for sure, they are certainly an advantage. Heat is probably one of the biggest issues, and so is speed and distance. light travels more efficiently the electricity so the gain would be good from point A to B. What I'm not sure about is the laser, they would have to make it quite small for it to fit in small chips. I assume they have that figured out already. |
| VictorEM83 - Feb-11-2004 server time |
| At first but isnt everything? Optics on chips is most likely the only way to get data off the chip in the future the pins will all be needed for power, due to SoI inability to stop leakage on the sub 90 nm processes. Optics wouldnt create heat like metal pins do cuz of resistance. Also Optics could lead to upgradable/modularl motherboards that users could configure or change as peoples needs do. Thats the amazing thing about optics, they could enable new things for PC that havent been feasible before. |
| Red Squirrel - Feb-11-2004 server time |
| Cool, but that would be quite expensive probably. |
| VictorEM83 - Feb-11-2004 server time |
| Intel uses silicon for 1-GHz optical modulator This is a start to having super fast fiber optic cables for HDD's and optical interlinks on motherboards |