To: MileHigh who wrote (8910 ) 10/25/1998 5:10:00 PM From: wily Read Replies (1) | Respond to of 93625
MileHigh: I got the answer back from the AMD thread, it only took a few minutes. Looks like the guy who wrote that article was on drugs or something. I'm going to ask the author if he cares to respond, but Maxwell's version has the stamp of reality. Wily: << "Intel is talking about [chips running at speeds of] about 1 gigahertz," Strauss said. Silicon germanium, however, "gets you up into the 50-gigahertz range," more than 100 times faster than Intel's current top-of-the line 450-megahertz chips. "This is literally a revolutionary increase in speeds," much more significant than IBM's new copper chip technology, Strauss said. While using silicon germanium in highly complex computer chips is several years away, IBM has a lead of at least nine to 12 months over competitors trying to develop silicon germanium technology, Strauss said.>> You seem to miss the point. 50GHz range is only for one single transistor switching speed. If you hook up 7.6M transistors together in a CPU like the PII you will be limited by your logic gate (a series of transistors hooked in series) propagation delay and RC time constant due to interconnect. The result is that the entire chip will run much slower than one transistor. For your information, Intel 0.25um transistor of 0.18um drawn gate on their PII-450 is currently the champ. This transistor can switch up to 30GHz. Motorola HiperMOS5 0.25um is in second with switching speed of 29GHz. AMD is approaching Motorola very rapidly. So don't listen to all the hype on GaAs and Germanium. Silicon still has a lot of leg room. It is up to the chip designers to squeeze out the speed. Maxwell Message 6149617