SoylentNews is people
SoylentNews
IBM has demonstrated a chip that can take advantage of photonics' higher bandwidth and lower energy consumption:
Engineers have long known that fibre-optic links are more desirable than copper wires for shuttling data around—the available bandwidth is higher, the distances that signals can be squirted over are longer, and energy consumption is lower. On the other hand, when it comes to actually doing stuff with that data, electronics are where it's at. This dichotomy has resulted in a very pronounced split between optical and electrical technologies: optics are used for networking between computers, but inside the chassis it's electronics all the way.
This approach has worked well so far, but as bandwidth and energy requirements continue to soar, research labs around the world have been looking at ways of bringing the optics ever closer to the electronics. The first step is to bring optical channels onto the motherboard, then onto the chip package, and ultimately onto the die so that electrical and optical pathways run side-by-side at a nanometer scale.
Quantum computing has also made a lot of gains recently. Perhaps in 5 years we'll be looking at a higher order of magnitude in processing power. What would you do with it?
(Score: 3, Interesting) by ikanreed on Thursday May 14 2015, @08:02PM
I think you accidentally mixed up "Quantum Computing" and "A magic wand".
The problems you've described (with the exception of the biochem one) consist entirely of unsolved engineering and scientific quanderies, not problems alleviated by more efficient computing power.
A quantum computer won't instantly make our weather models have more precise inputs, or a better underlying model of weather patterns. Research meteorologists have a lot of work to do that a computer chip can't replace.
Nuclear fusion is well understood conceptually, but requires a lot of meticulous careful design of machinery to induce it, as well as possibly some research breakthroughs in the behaviors of high-energy plasmas.
Human brain interfaces aren't short of processing power, but instead need a lot of experimental observation to determine what easily detectable brain activity is meaningful.
The list goes on like that. Quantum chips reduce the complexity of large-space algorithms, they don't magically solve problems for you.