weird restarts when compiling

[Chuck Swiger]

On Jul 13, 2008, at 4:38 AM, Aggelidis Nikos wrote:
> FATAL ERROR: Rounding was 0.5, expected less than 0.4
> Hardware failure detected, consult stress.txt file.
> Torture Test ran 9 minutes - 1 errors, 0 warnings.
>
> so maybe it isn't even a psu problem but a cpu one?

If it was strictly a problem with the CPU, you'd likely not be able to  
boot, or this test would fail within a few seconds.

9 minutes is possible but a bit on the early side to be a thermal  
issue; more likely would be the memory config and BIOS setup is  
marginal and that relaxing the timings would help.  If you shut down  
for a while, and then do a cold restart with the case open, does that  
result in the stress test taking longer to fail?

If so, check your cooling.

On the other hand, if the length of time before failure doesn't  
change, or more precisely seems random-- 5-10 minutes one try, an hour  
the next, that generally indicates memory and motherboard BIOS setup.

PSU problems tend to not show up with prime95, but when you have the  
disks busy and maybe graphics, also.

-- 
-Chuck