home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.arch
- Path: sparky!uunet!cis.ohio-state.edu!zaphod.mps.ohio-state.edu!usc!sdd.hp.com!news.cs.indiana.edu!umn.edu!csus.edu!netcom.com!tcmay
- From: tcmay@netcom.com (Timothy C. May)
- Subject: Re: Multi uP P5 systems
- Message-ID: <ttynrjr.tcmay@netcom.com>
- Date: Mon, 14 Sep 92 17:06:27 GMT
- Organization: Netcom - Online Communication Services (408 241-9760 guest)
- References: <1992Sep14.143344.1051@crd.ge.com>
- Lines: 49
-
- davidsen@ariel.crd.GE.COM (william E Davidsen) writes:
- : | On a tour of a major MB manufacturing plant my friend showed me what they
- : | were working on in the design lab. It was an EISA board with sockets for
- : | four P5 chips! He said the P5 has features that make it very easy to
- : | design multi-CPU boards.
- :
- : I hate to say this, but what is anyone going to plug into the sockets?
- :
- : Rumor of the day: Intel has dropped the P5 design as unmanufacturable
- : and is redesigning the chip in 4000 angstrom biCMOS to improve yeild.
- : They will never sell commercial quantities of the P5 as it has been
- : shown.
- :
- : The Intel presentation at FORUM seemed to hint at this, and I've heard
- : other things, would someone from Intel care to deny this?
-
- I think a more plausible reason for the delay was the sudden
- disappearance of AMD as a short-term threat to the 486 pricing
- structure. This gives Intel several additional months, at least, to
- sell 486s at a higher price than would be likely if the P5/586 was
- shipping. (Cyrix may be a more realistic threat, but their full-blown
- 486 clone is not yet announced, is it? I'm unsure about their situation.)
-
- And the several extra months can always be used to ramp up volumes,
- further test and tweak the design ("tweak" as in minor steppings,
- _not_ major redesigns). Without a lot of market pressure to announce
- (and what some users _want_ is not the same thing as real market
- pressure), why should Intel announce?
-
- The 3.5 million transistor P5 is, from reports, built on the same
- process as the 3 million transistor 860XP, so presumably Intel has had
- an estimate of expected yields for at least a year.
-
- In any case, the announcement is still reported to be coming soon, so
- we'll all know the truth soon enough. (This also goes against the
- "complete redesign" rumor, as such a redesign would mean a _much_
- longer delay. Finally, these things have a way of reaching the stock
- market very quickly, to wit, if the P5 was in real trouble, I doubt
- Intel stock would be near its all time high, at $65+)
-
- --
- ..........................................................................
- Timothy C. May | Crypto Anarchy: encryption, digital money,
- tcmay@netcom.com | anonymous networks, digital pseudonyms, zero
- 408-688-5409 | knowledge, reputations, information markets,
- W.A.S.T.E.: Aptos, CA | black markets, collapse of governments.
- Higher Power: 2^756839 | RSA MailSafe Public Key: by arrangement
-
-
-