Talk:Advanced Computing Environment

Untitled
ACE was supposed to provide full parallel support for x86. MIPS machines were to be the "high-performance", x86 the high-volume end. Internal politics (like DEC wanting to go with OSF/1 for Unix instead of SVR4) and technical problems (like the R4000 shipping later than expected) delayed products and contributed to the failure of ACE. -Pelladon 07:26, 10 June 2006 (UTC)

Removed self-declared unsubstantiated rumor
I pulled the following sentence from the article:

One of the first companies to leave the consortium was Compaq. The unsubstantiated rumor was that Intel threatened that company to not deliver sufficient quantities of processor chips to support Compaq's then current production rate. (emphasis original author's)

If there were at least an citation for this (a sourced unsubstantiated rumor? :-), it might be OK to leave in, but with nothing to back it up at all, it didn't seem appropriate.--NapoliRoma 18:04, 3 May 2007 (UTC)

RISC vs. CISC performance claims
As Rilak has pointed out, there are problems with the following: When the initiative started, RISC based systems (running at 100-200 MHz at the time) had substantial performance advantage over Intel 80486 and original Pentium chips (running at approximately 60 MHz at the time). Intel quickly migrated the Pentium design to newer semiconductor process generations and that performance (and operating frequency) advantage slipped away.

First of all, the MHz values for that time seem off. Second, megahertz is a completely bogus metric for performance comparisons, especially between unlike architectures. What would be meaningful here would be something like SPECint and SPECfp values (and prices!) for comparable systems of the day, 80486 vs. MIPS, since that's what ACE was targetting. I believe floating point was a major weakness of x86 systems at that time, which would be a critical factor in the workstation market.--NapoliRoma (talk) 19:41, 3 July 2009 (UTC)


 * After all this time, I recently modified the first part of this and dropped the "original Pentium chips" part, along with the note about Intel's subsequent Pentium developments. ACE was around before the Pentium reached the market. I imagine the person who wrote this section didn't get their timeline right. PaulBoddie (talk) 21:18, 11 October 2022 (UTC)