September 5, 2001

9:36 AM

Motorola in the fast lane. Motorola researchers in Tempe have developed a semiconductor that runs 35 times faster than today's models. The research has solved a 30-year-old problem of creating a semiconductor that combines the durability and economy of silicon with the high speed of crystal compounds used in lasers and fiber optic applications. The new wafers will be licensed next year, but the company doesn't expect to see products on the market for another two years. The semiconductor runs at 70 gigahertz instead of the current 2 gigahertz, the speed of the fastest processors in personal computers.
posted by 120degrees (5 comments total)

Alas, we won't be seeing this in CPUs anytime soon. (nor in any chips until 2003 or so) Especially not with Motorola focusing more on the embedded market.
posted by darukaru at 2:28 PM on September 5, 2001

Apple: "Uh MHZ myth? Did we really say that? Well we were kidding. Faster is better now we REALLY are ahead of Intel."
posted by geoff. at 2:40 PM on September 5, 2001

If it's used for RAM, it could virtually eliminate the complicated cache architecture modern processors require to keep them fed with data.
posted by kindall at 2:44 PM on September 5, 2001

Unless I'm thinking of the wrong press release, this is basically a way to get cheap gallium-arsenide wafers, and will be used to cut costs on embedded/RF gear, not for speed increases.
posted by aramaic at 3:36 PM on September 5, 2001

It's easy to misunderstand this technology. Aramaic is right: this is a technology used for RF and optoelectronic devices. I don't believel that GaAs technology (whether on a GaAs substrate or on a silicon substrate) is currently capable of achieving the kinds of densities and yields which would be needed to produce microprocessors or RAM or similar devices.

I've seen this reported in a number of places now and the reports of the claims conflict pretty substantially. Generally, the less technically-sophisticated the publication, the more enthusiastic the claims seem to be. The report which seems to be the best was EE Times, which not surprisingly was written by someone who actually understood what was going on.

There's much less to this than meets the eye. And no, this isn't going to dig Apple out of the hole it's in.
posted by Steven Den Beste at 4:09 PM on September 5, 2001