How long to torture test?

[kh]

I am currrently waiting for my XP 1600+ chip to get here from newegg so I decided to mess around with my current chip which is a locked 1.33ghz t-bird. I 've always had my fsb at 140 (to be safe) and never had a single problem on my shuttle ak31a (rev 3.1). I want to to find out the max fsb my components can handle and to do this I am using prime95 as a test. My question is how long (in general) is a good length of time to let it run so I know that everything is stable? Thanks for any help!

EDIT:

Forgot to add: What is the highest I should ever go with the Vcore? It was at 1.75 when I was at 140 fsb. Thanks again!

 

[Thor86]

Run Prime95 for at least 8+ hours. That should be good enough for any stability- related tests.

 

[SupermanCK]

depends on your need...if you are planning on running certain programs for however long...then test that long plus a few...
if you are only running short programs...just test for 6+hours

 

[MoFunk]

Originally posted by: kh
My question is how long (in general) is a good length of time to let it run so I know that everything is stable?

Make sure you run Seti on it for the TeAm 24/7. That will burn her in real good! Check us out

 

[BuddyAtBzboyz]

Umm he's not burning it in, he just wants to make sure its stable. However if you are going to be running seti on the computer, you might wanna test with prime 95 for 24 hours.

 

[SupermanCK]

you should not go higher than 1.75 in windows reading idle or under load.....
usually under load's vcore is lower...

 

[RaynorWolfcastle]

Originally posted by: SupermanCK
you should not go higher than 1.75 in windows reading idle or under load.....
usually under load's vcore is lower...

is he talking about overvolting or just core fluctuation?

If you're talking about Vcore fluctuations, they should be very small

OTOH If your talking about overvolting, 1.75 is default. overvolting by =< .2V is usually considered accepteble (Proceed at own risk)

-Ice