Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Intel Corporation (INTC) -- Ignore unavailable to you. Want to Upgrade?

To: Tenchusatsu who wrote (72331)	1/28/1999 2:15:00 AM
From: Paul Engel	Read Replies (1) \| Respond to of 186894

Tenchusatsu & Intel Investors - Intel and RAMBUS are getting CLOSE ! Toshiba has just delivered a "B" version of Direct RDRAM RIMM Modules to Intel for full speed 800 MHz testing. Any bets as to what Intel will be doing with these ? Pentium III isn't due out for another month - on 440 BX chip sets - and Intel is way down the road on the Camino/820/Rambus/133 MHz/800 MHz upgrade. Who is the guy on the AMD thread who keeps complaining about Intel's aggressive technology introductions? Paul {=================================} semibiznews.com A service of Semiconductor Business News, CMP Media Inc. Story posted 7 p.m./4 p.m., PST, 1/27/99 Toshiba ships Rambus RIMMs to Intel for 800-MHz testing IRVINE, Calif. -- Toshiba America Electronic Components Inc. here today announced it has delivered samples of 64-megabyte and 128-Mbyte Rambus RIMM memory modules to Intel Corp. for testing in high-speed systems. The U.S. subsidiary of Japan's Toshiba Corp. claimed the company has become one of the first suppliers to provide samples of revision B Rambus modules that allows Intel to test fully loaded 800-MHz systems with 32 RDRAM devices, based on the memory architecture from Rambus Inc., Mountain View, Calif. "The process of going from the Direct RDRAM design to mass production involves several steps," said Stephen D. Marlow, vice president of business development in the Memory Business Unit of the Toshiba subsidiary. He credited Toshiba's Scalable by Design program as enabling the company to "to hit every milestone. Toshiba said its RIMM modules were demonstrated in test boards running at 800-MHz at the developer forums hosted by Intel and Rambus last September.

To: Tenchusatsu who wrote (72331)	1/28/1999 10:50:00 AM
From: Scumbria	Respond to of 186894

Ten, The problem that K7 faces with it's architectural excesses are two-fold. And both are very serious problems. 1. The large die size increases defect/die probability and decreases die/wafer, resulting in significantly lower yields (and higher costs.) 2. The large die size also requires longer wires to route critical path control and data, resulting in lower clock speeds. AMD needs high MHz, low cost CPU's. There clearly has been some emphasis on maximizing K7 MHz, but the architectural excesses are unwarranted and self-defeating. At this point it appears that Dirk Meyer made a lot of the same mistakes on K7 that he did on the 21264. I'm not saying that K7 will flop, just that AMD could have made their path much straighter and simpler. Scumbria

To: Tenchusatsu who wrote (72331)	1/28/1999 10:54:00 AM
From: Tony Viola	Read Replies (1) \| Respond to of 186894

Tenchusatsu, >>>Thanks, Scumbria. I guess you're confirming what I've suspected all along, that the K7 might be overdoing it on execution units and buffer sizes. But like I said before, it's not a bad idea if you have a sizable transistor budget to work with in the first place.<<< Yes, but a company that has a history of being plagued with yield problems maybe should be extremely judicious with that budget. Tony Editing...As I catch up reading, I see that Scumbria said exactly this in post #72365. There clearly has been some emphasis on maximizing K7 MHz, but the architectural excesses are unwarranted..<<<