Error during Prime 95 torture test. Is it more likely that it's due to the total Mhz or high fsb?

[Parn]

I am running athlon xp 2000+ at 166mhz fsb x 10.5 = 1743MHz

I get errors in Prime95 torture test. Do you think it is more likely this is due to the high fsb or the high cpu speed?

This is the error I got:

FATAL ERROR: Resulting sum was 986718960.7295277, expected : 986910151.6938999
Hardware failure detected, consult stress.txt file
Torture test ran 1hrs, 15 minuntes - 1 errors, 0 warnings

Athlon XP 2000+ AGOIA @ 166fsb x 10.5.
Crucial Micron 256mb PC2700 RAM
Epox 8k3ae (has 1/5 divider)

 

[Duvie]

What is the mobo and does it have a 1/5 pci divider??? If it only has 1/4 divider then the pci bus speed is 41.5mhz which is high....

Also you are only 80mhz over stock speed so the total does not seem excessive...Have you tried bumping the vcore sligthly to see if it helps???

The high fsb would likley be more of a problem for the chipset of mobo, pci and agp devices if no 1/5 divider, and memory if you are running a high ratio.

Using a 1:1 ratio now would be 333mhz or pc2700ddr...Do you have pc2700 rated ram??? If not try setting ram timings less aggressive and try again...

Also list error...Is it rounding error or hardware error???

 

[First]

More often than not, Prime95 failures seem to be a result of some type of memory setting that the motherboard can't handle (whether it's too high a frequency, very aggressive timings, whatever).

But list what Duvie asks, since that'll be the only way anyone can accurately diagnose your problem.

 

[Parn]

Edited the post to show what I have

 

[jaybee]

What are your memory timings (latencies)? What is your vcore (cpu voltage)? Can you complete a 3dmark? My guess is you need a little more voltage to the cpu. What are your load temps like? LMK about my ?'s and try bumping vcore .025-.05v -- I bet that'll do the trick.

jaybee

 

[Parn]

edited with error message

 

[Duvie]

Hmmm....

IMO...When on gets a rounding error it has generally in the past for me been a vcore issue (more power) or just to much of an overclock. However IMO again hardware errors like that one have been a result of overclocked memory, too aggressive of timings, and/or bad stick of ram....

Couple of things to do....

1) Run memtest on the ram to make sure there are no errors.....

2) Take ram stick out if you have multiples and run one at a time and try maybe each bank if you can.