Vcore/cpu experiment

[work permit]

I ran an experiment to see what the tradeoff is between higher vcore & higher cpu temp. Running sandra cpu benchmark in a loop, and monitoring cpu temp in mbm, I slowly increased cpu temp by shuting off my cpu fan & occaisonaly turning it back on @ 7 volts. OC'ing a 1 Ghz piii, the cpu temps that crased were:

FSB 1.75v 1.85v 1.95v
147 50 61 68
148 48 57 64
149 45 55 60
150 42 52 57

Motherboard temp was 28-29 in all cases.

I understand the above temps don't represent a fully stable system. I'm hoping this gives me a sense of how much to raise vcore after testing for stability to give some "head room" for changes in ambient temp.

[FlippedBit]

Thanks for posting your data. If you get any more results please post them too. Much steeper gradients on your temp versus fsb and vcore surface than I would have guessed. Very interesting.

[Torstein]

Yes, very interesting experiment. Would be interesting to see if you were able to get consistent results when if or if you repeat the test runs under same conditions.
However, I think you are conducting somewhat of a type of destructive testing to the benefit of us all.
In the area of oc-experimenting it should be interesting to see how many are experiencing thermal runaway and under what conditions?

---------------
CUSL2 PIII700e@1001 Sandra mem bm 418/502
Vitelec 2x128_143 at 222-5/7
temp 24/41 degr.
FOP32 HSF

[work permit]

i've spot run a few points and do fet similar results. don't want to repeat entire experiment...willing to shave a few minths off processor life but not years

also, i only ran a narrow range of fsb...the range i'm interested in