To: Steve Lewis who wrote (7416 ) 3/17/1998 12:17:00 PM From: Rob S. Respond to of 11555
Let's play Devil's advocate: "So IDT has an agreement with IBM, so what? Cyrix has been manufacturing their parts at IBM and that didn't protect them from seeing losses in the past. AMD also has an agreement with IBM who will begin making their K6+ soon. So what's so special about the agreement between IBM and IDTI?" Both Cyrix and AMD's tactic in going after Intel was primarily to build chips along the same "bigger and better" uP design strategy that has proven to be counter productive for Intel, AMD and Cyrix. This resulted results in very large, complex, hard to test and power hungry chips that end up consuming more wafer real estate and have lousy yields. AMD has been stung badly by poor yields on the K6. Cyrix lost money and was acquired by NSM. Glen Henry's, (Centaur/IDT), tactic was to take a fresh look at the uP architecture, simplify it as much as posible and also take advantage of the capability of putting larger and more available memory on the chip. "K eep it S imple S tupid" rather than continue to pursue the now out of step policy (for the popular PC market) of trying to pack more performance into the uP while the real limitations have more to do with I/O than raw, theoretical processing power. True genious is often disguised in simplicity. Cyrix and AMD are or soon will be producing uPs at IBM. But both of these guys have failed to produce highly manufacturable parts. Both of them have had lousy yields and recent losses. IDTI has already been reported to be producing parts in their similar facilities with high yields. Garbage in garbage out, or in IDT's case, choice currency paper and ink in, bankable currency out. That is the only equation that realy matters in the IC industry and one in which IDT/Centaur has a clear advantage over competitors.