Next on the research list is to integrate the lasers with waveguides and drive electronics
I always thought that the shark is the traditional first integration subject.
Silicon photonics boosted with UK fabrication research
Silicon photonics is one of the industry's hottest research fields, because it holds out the promise of accelerating on-chip communications without the extra heat that faster copper-based comms generate. One of the big challenges is fabrication, and that's the subject of research announced by University College London, Cardiff …
-
-
Thursday 17th March 2016 08:24 GMT Dave Pickles
1300 nanometre?
Presumably the laser has to be at least as big as the wavelength of light it's generating, and the receiver must be a similar size? With semiconductor geometries now down to 10nm you're going to have to forego an awful lot of transistors just to send an optical signal across the chip. A big fat copper conductor would be more economical in area terms.
-
-
Thursday 17th March 2016 09:32 GMT Andy E
Re: 10 Years only?
I was thinking the same thing. Do we now have to start tracking the deployment life of the CPU's so we can plan for the expected failure after 10 years?
The Reg ran a series of articles on very old equipment that was still in active use because it continued to do the job. In most cases it was impractical to replace the kit due to lack of working knowledge or modern equipment being unsuitable.
In the early days the performance of CPU's regularly doubled as the clock cycles got quicker, the on board caches got bigger and they became more efficient. This seems to have calmed down with new CPU's becoming less power hungry but offering more core's, better efficiency and better support for VM. This will probably mean that the usefulness of older CPU's will be extended as it will be hard to justify the upgrades on a cost/benefit approach. 10 years expected lifespan might become 20 years.....
I'll get my coat. It's the one with the ZX Spectrum in the pocket.
Andy
-
Thursday 17th March 2016 12:43 GMT allthecoolshortnamesweretaken
Re: 10 Years only?
So you wouldn't put it on a space probe. Plenty of other applications where 10 years is just fine. And 10 years ist just for starters anyway.
The bigger question is, will this actually lead to some widget or other that will be produced in the UK?
(Pro tip: bulid something looking really cool for the minister to stand in front of. Doesn't really have to be related that close to the actual product. Or claim that it could be weaponized, maybe. Remember: "Funding makes this thing go up. No bucks, no Buck Rogers.")
-
Biting the hand that feeds IT
