Laser microchips set to speed computing

Great, I can't wait to see what games NEED to have that type of processor.

actually, Sony has been making many significant breakthroughs in that field and recently found out how to create fiber optic interconnects for systems.

I bet Devil May Cry 5, with real-time environment changing, will require laser microchips along with Blu-Ray storage capacity.

oohhh.. Lasers!! Maybe we'll have DNA based harddrives, wireless power supply, 3D monitors and Blue-Ray by then as well. ok, maybe not wireless power supply i guess.

It's funny to see how long it takes news like this to trickle out to the mainstream. It has been what? Over seven years since I first heard about this research.

It took years to finally get dual core processors and dual/ multi-processor systems in the hands of the general public, so I wonder how long it will be before this becomes commonplace technology?

AMD, Intel and IBM have invested alot of research into the paradigms they are currently heading down: 64bit, multi-core x86, and in IBM's case networked cells. It's in their best interest to milk their current path and continue "researching" optical computing.

When processor growth becomes stagnant and the major companies need to feed Moore's Law http://en.wikipedia.org/wiki/Moore's_law , you'll start reading press releases about the big three supplying OEM manufacturers like Dell and Lenovo with them the very next quarter.

http://en.wikipedia.org/wiki/Moore's_law#Future_trends

Bring on the sharks!

"There's a lightshow in my processor!"

I'm personally enjoying this idea in terms of how it will evade "Oh noez! Wires are touching!"

soon we'll see atom chips that harness the power of atoms.

Photonic computing has been "around" for ages. I remember reading about it alongside "quantum computing," which (theoretically) would be even faster than using lasers. Faster than light?! Yes, accessing superposed (and common) states would be quicker than moving light around your computer's innards.

In any event, successful implementation of this stuff on a mass scale is still ages away, and given the pitfalls (and there are many), this could go down along Duke Nukem as the next big vaporware.

so much for the Cell processor and its development costs.
if the claims by UC Santa Barbara are correct, this technology would pretty much render the concept behind Cell processors useless.

*a la Nelson* ha-ha!

I wonder what happened to DNA computing? Ten years back, Time and Newsweek were discussing canisters of genetic material which could be used as super high-capacity hard drives. Perhaps scientists just decided that organic material was too unreliable and scrapped the idea.

Light-speed processors are a very exciting prospect, however. Computers with this technology will not only run quickly, but much cooler than traditional PCs. Just imagine a laptop that doesn't bake your cajones after using it for a couple of hours!

JR

I might be missing something but isn't the speed of electricity 1/3 of light? How do you get more than 100x?

SWEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEET another tight tech, us geeks/gamers shall one day be one of the most skilled on earth at everything. Pretty much guranteed, I mean, so many e-news that has been talking about stuff like "car games help real life driving skills" and "1st person shooter games helping people with aiming skills" stuff like that, pretty cool stuff.

Silicon microprocessors are going to have to become obsolete sooner or later, I'd prefer to see it sooner. Hopefully not in 2024 like Dr. Who predicts, but in 2010.

jc -

The current computer designs have a problem with resistances. Electricity in an optimal state may move 1/3 the speed of light, but when you start throwing in resistors and all the conductors, it loses a lot of speed. The laser design moves light through a vacuum, which eliminates any speed loss.

ManekiNeko -

DNA computing is a pain in the ass. They have to deal with a 0-3 (one for each type of molecule in a DNA strand) instead of binary. Basically programming would have to be rewritten from scratch. From what I've read about DNA Computing is that all the extra work would only result in speed increases, but does not gain any benefits in efficiency under the computational complexity theory.

Now, Quantum Processing, that sounds tasty.

This is all a lot of hubbub really. Until someone actually produces a tech and makes it work there is VERY good chance that it will never see the light of day.

Quantum computing was a big fuss a few years ago, todate nobody can really get it to work. Now other forms of technology have caught up to it to the point where most experts think the only real application for a QC will be in high end encryption. For every other application they will just be too difficult to program for.

Optical based processors potentially can run a lot faster than just about anything else around, but contrary to the thought that these run cool, the power density in an optical processor is enormous and is a big reason why they are having trouble getting the tech up and running, let alone at acceptable performance levels required to displace the current tech.

the other point is that to truly get the speed out of an optical system, you need an entirely optical system. having ANY electrical components produces bottlenecks and severly limits potential speed, which means optical fibres and junctions running between EVERY component in an entire computing system. To do that at the moment you would have to build a computer 10 times the size of your regular PC tower, at least.

As I said, don't get too excited for revolutionary new tech till someone shows you it working.