CPU FSB Mem PCI
1900 350 350 35 - Great for 2 months then bluescreens...
1666 333 333 33
1500 300 300 30 - still bad (!)
Memory tested extensively as not the problem,
Graphic card appeared fine.
Computer still crashed with L1&2 caches turned off,
(despite running at 1/100th speed)
Memory benchmarks and little things, produced unhandleable
bluecreen errors for PRN blah, ntoskrnl, mup.sys -allsorts
At last its running stable at
CPU FSB Mem PCI
1320 240 306 30
70% of it previous stable speed, least its stable again.
What did it? -i think, a bit off superglue on the cover of
the southbridge,i forgot about- a few days before the crashing began,
its currently reporting a temp of 39C
I thought it could be Power supply since slowing everything could cure it
-The southbridge heatsink (or lack of), I hope thats the problem.
' optimising, accomodating bodges,
I was using Corsair Value Select PC2700 that I ran at agressive timings
2-2-2-6 (stock timings were 2.5-3-3-7) and somewhere between 340 and 350 FSB
for probably a year off and on. At some point it became too much so I had to
loosen up the timings to stock. Granted I probably had no business running
at those speeds and timings but there you have it. So it would appear that
one can wear memory out at some point.
I'm running OCZ Performance PC3200 now but I can't realize the full FSB
speeds (200mhz) on this motherboard. It accepts the DDR400 ram but since
there is not ratio setting or AGP lock it's running about the same as
before. But that's a question for another thread.
"Creeping Stone" <l...@l.lll> wrote in message