Daily Archives: July 20th, 2007

Friday Science

I’ve been a computer user for ten years now. Since I started in my late thirties, I was way behind the curve when it came to knowing how to use a computer and utilize the various programs involved. To say the least, I’m quite an accomplished Microsoft Office user now (I found out later that’s not a claim to fame). Anyway, over the years I have taken various computer courses, either in programming (C++ and Visual Basic, I suck) , HTML (not very good at either) and more updated Microsoft stuff (no Vista yet). I’ve even taken a hardware course, so I know some basic stuff about processors, buses, chips, ROM, RAM, bios, clockspeed, etc.

But when it came to chip and processor manufacturing, I knew diddly squat. I knew most chips were ceramic, with copper, silver, nichol and even gold inlaid into them. And the more pathways squeezed onto the chip, the faster the processor speed. But the trouble with that is that the heat generated is higher as the speed increases. Big problem. It also makes it looks like there’s a limit to Moore’s Law.

Enter Optical Processing. Optical transistors have been in the works for a while now and even though the processing speed is greater (speed of light) the energy required to use them is quite high. The reason is that these transistors has to utilize a whole beam of light to activate them from one state to another. But now researchers have found a way for the transistors to activate using a single photon of light:

Mikhail Lukin and colleagues at Harvard University have come up with a technique that uses a single photon to switch the state of a light beam. This is the first workable suggestion for building an optical computer, they say.

An electrical transistor’s speed is limited by the speed at which an electric current flows. In theory, because photons travel much faster than an electric current, substituting photons for electrons would speed things up. In reality, however, finding the optical equivalent of a transistor has proved difficult.

…Lukin suggests using a semiconducting nanowire that resembles a miniature optical fibre, and sending both beams down it. Because the nanowire confines the rippling plasmon to a smaller space than the gold film, it should be more sensitive to changes in the intensity of the control beam. Lukin says that this increased sensitivity is enough to allow just a single photon to switch the state of the signal beam.

Nanotech is going to play a huge part in the next stage of computer evolution, just by reducing the size of the carrier fiber, light can be taken down to the quantum level photon that can actuate the transistors. This reduces the energy requirement and waste heat also, two for one!

I wish I was smart enough to get a job doing this stuff!

(Speed of Light Computing Comes a Step Closer)