Microsoft is Screwing CYRIX on behalf of Intel
Paul
Read this fellas,
I found this information very enlightening and I think you will to,
judge for yourselves.
"In testing the final release candidate of Microsoft Windows NT
Workstation 4.0, BYTE has discovered a performance discrepancy when
running the software on a
Cyrix 6x86-based PC. While running the Bapco SYSmark/32 suite of
32-bit Windows applications on a Cyrix PC, BYTE author Rob Hummel
noticed a
slowdown of approximately 25 percent with the final release candidate
of NT 4.0 (compared to an older beta version). In both cases, BYTE
tested NT on an
identically configured PC that uses a Cyrix 6x86 processor running at
150 MHz. That processor is known as the 6x86-P200+.
Pentium-based PCs running the same NT beta software did not exhibit a
performance decrease from one NT beta to the next. Additionally, NT
4.0 running on
Pentium PCs ran the SYSmark/32 application suite an average of 30
percent faster than with NT 3.51, and 12 percent faster than Windows
95, on the same
Pentium system. Pentium Pros ran NT 4.0 26 percent faster than NT 3.51
and 28 percent faster than Win 95. The Cyrix 6x86, on the other hand,
ran NT 4.0 16
percent slower than NT 3.51 and 24 percent slower than Win 95.
Microsoft added code to NT to make it more stable when running on
Cyrix-based PCs, confirmed Phil Holden, a product manager in
Microsoft's Desktop &
Business Systems Division. This may be the source of the application
performance slowdown. But Microsoft officials declined to comment
further, saying only that
the company will have more information on the matter within the next
few days. Cyrix officials told BYTE they are currently investigating
the matter."
Microsoft "ADDED CODE to NT" under the pretense of making it more
"STABLE" with Cyrix processors. In other words, they added special
code which detected Cyrix processor that disabled write back cache
ability of Cyrix chip and allowed only write-through so in once simple
step cutting the balls of Cyrix processors and eliminating the
competition for Pentium processors in the very equitable high end PC
market.
These guys are vicious, no denying, this is for all those wanting to
know about a so called "BUG" in the Cyrix. Sounds more to me like a
"BUG" intentionally placed in NT 4.0.
Paul(por...@onramp.net)
Jason J Ellingson
You are late in pointing this out. Also, you are somewhat misled. That
slowdown only affects the Rev2.6 or older versions of the 6x86. Cyrix
acknowledges this and will send you a replacement chip (with proof of NT
purchase) to correct that problem. Revision 2.7 and the newest 3.7 do not
exhibit this problem and run at full speed.
So please stop jumping at Microsoft. They may have saved some business a
million dollars from faulty data going through their NT server!
--
Jason J Ellingson
elli...@execpc.com
Paul <por...@onramp.net> wrote in article
<327253b3...@news.onramp.net>...
ji...@hal-pc.org
I> These guys are vicious, no denying, this is for all those wanting to
> know about a so called "BUG" in the Cyrix. Sounds more to me like a
> "BUG" intentionally placed in NT 4.0.
> Paul(por...@onramp.net)
>
Maybe I'm naive,but I don't think so. Reason : It sounds like a very lucrative lawsuit for Cyrix if true.
-----------------------------------------------------------------------
This article was posted to Usenet via the Posting Service at Deja News:
http://www.dejanews.com/ [Search, Post, and Read Usenet News!]