Martin wrote:

>> Running 3 threads of "small FFT test" works fine, nothing unusual to
>> report.
>> Running 3 threads of "large FFT test" makes the system hang after a
>> couple of minutes.
>> Running 3 threads of "blend" makes the system reboot after
>> 30-60seconds.
> 
> OK, the interesting bit is whether you get those same results
> independent of _which_ core you disable.
> 
> Can you test for when you have core #0 disabled, then with core #1
> disabled and so on?
> 
> The suspicions are that you might see core #2 fail sooner than the
> others...

Ah, I see - we're looking for (possibly) bad core.  Yep, I'll do the
tests again with different cores off-line. 

> I think the most useful tests will be when you have a known good
> heatsink contact!

I'll re-seat the whole thing before I start.  What you and David
suggested about pockets of air between the CPU and the heatsink makes
perfect sense. 



/Per Jessen, Zürich

_______________________________________________
Prime mailing list
[email protected]
http://hogranch.com/mailman/listinfo/prime

Reply via email to