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.

Politics : Formerly About Advanced Micro Devices -- Ignore unavailable to you. Want to Upgrade?

To: RDM who wrote (66695)	7/26/1999 7:58:00 PM
From: Ali Chen	Read Replies (1) \| Respond to of 1573954

RDM, <A Concise Review> Thanks, the article seems very good. I think that every paragraph of this Review should be enumerated sequentially, so we could easily communicate any proper part to thread "theorists" like Elmer, Engel, and Tench when pointing to flaws in their barking. Of course, the article is politically polite, so it requires some thinking when fitting this science into practicalities of x86 arena. For example, I found this interesting: "Transmission-line problems with shared multidrop buses, however, preclude operation much above 133 MHz. Next-generation buses, however, will use point-to-point signaling with source-synchronous clocking,.." Therefore, K7 seems to be classified as "next-generation" while the P-III/X-III/... uses apparently an old obsolete technology... BTW, the RAMBUS signalling appears to be somewhat eclectic - it must be "multidrop" for system expandability, but must use a sort of clock forwarding to get into 800MHz signal rates. I see some "difficulties" when implementing this bus in the field... For such a short article, there must be some ambiguities. For example: "Lengthening the pipeline, or superpipelining, divides instruction execution into more stages, each with a shorter cycle time; it does not, in general, shorten the execution time of instructions. In fact, it may increase execution time because stages rarely divide evenly and the frequency is set by the longest stage." For some semi-literate minds like Elmer Youseless, it may be hard to understand the meaning of "is set by longest stage" - they may fall into a terminological discussion about how sequentially connected gates are different from combinatorial logic. Regards, - Ali