I just installed my Cyrix 6x86MX-PR200 in an Abit PX5 motherboard with 32 megs
SDRAM and the system is running slower than my old Cyrix 6x86-PR166+/MTech
R533/32 FPM. I don't understand it. Running Sysinfo for DOS, I got 1024 CPU
mark for the old system and 990 with the new one. I have it running at 83*2
whereas the old system was at 68*2.
It seems faster, but even in GP2, the game play is much slower, with Processor
Occupancy well ove 115% at Spa where before it was below 100%. Does anybody
have any more suggestions to help me out or should I return it for another
chip?
Thanks...
--
Proud owner of a USR X2 v.everything modem.
--
*****The WINGMAN*****
>Something definitely isn't quite right. Cyrix recommends that the
>PR200+ is run at 75*2 (150mhz) which should yield an SI V8.0 of
>around 1020.
I forgot to mention that the SI of 1020 was with the M1 processor.
I don't know what it should be for the M2.
Mike Radler
Hawaii Ace League http://www.dithots.org/hal
>According to the PX5 specs on their website, Abit does not claim
>compatibility with the Cyrix 200. Here's a quote:
Abit does claim to support the Cyrix M2 with the latest PX5
(w/bios 3U). The M2s are the 6x86MX PR166, 200, 233). If
you support one you should be able to support all 3 as long
as the board can handle the clock multipiers, bus speeds and
voltages required by the M2 which the PX5 definitely does.
It also supports the new 83mhz bus speed. Having said that,
I checked Tom's hardware page and he found that the PX5
performed ok but was unstable with the PR233 especially with
the higher bus speeds, Hmm.... You can see his test results at:
http://sysdoc.pair.com/motherboardnews797.html
Cyrix is also recalling some 10,000 6x86MX PR200s due to a high
failure rate. I don't know the specifics but here's some info:
http://www.zdnet.com/zdnn/content/pcwo/0827/pcwo0007.html
Mike
Hawaii Ace League http://www.dithots.org/hal
Well that could be your problem right there. Remember SI has been
around for a long time, before the M1 and M2 were. Also it doesnt take
much to get SI to start giving dodgey feedback. It's just not an
accurate way to test cpu speed. I'm not really argueing against
anything else except the validity of using SI in these more modern
times.
Let's see: The old 686L PR166+ was really 2x66=133 (I got 899 on Norton
8.0 SI)
and the old 686L PR200+ was really 2x75=150 (I got 1013).
Now the new 686M2 PR166+ is really 2.5x60=150
and the new PR200+ is 2.5x66=166.
They've reduced the bus speed from 66 and 75 down to 60 and 66. That makes
the chip slower in some applications. Certainly, 2x83 would be faster than
2.5x66.
My old 686L overclocked to 2x83 gave a Norton 8.0 SI of 1125. So your 1024
does sound a bit low.
TRY SPEEDRATE FROM ZDNET...Better than SI.
