superpi 1M fails where prime95 passes?

[ArchAnGeL]

I am looking at my 3500 under phase cooling and while it primes away like a good 'un, if I try the superpi 1M I get the "non convergent in sq" error.

I've never know this before, always considering the 1M a pretty tame test of stability. If had very "on the edge" overclocks pass this before while falling down in other tests, especially prime95.

any ideas?

 

[Drazic]

Could be your cpu is cold buged

Your hd on the way ut or your widnows is

Don't run super pi when priming

 

[ted34]

Were you doing small tests in prime as that tests cpu and very little memory whereas superpi tests memory as well so could be your memory at fault

 

[ArchAnGeL]

Could be your cpu is cold buged

Your hd on the way ut or your widnows is

Don't run super pi when priming

Sorry, I wan't clear. This is when running superpi on it's own.

Perhaps cold bugged as it was running superpi fine on the same insatllation and hardware before I installed the mach 2. Don't suppose you have a link to save me searching around?

 

[welshtom]

Whats the stepping? What core is it?

 

[ArchAnGeL]

Whats the stepping? What core is it?

Winchester

 

[welshtom]

Quite a lot of winnies didnt like the cold.

run the cpu hot (with some load program) then run super pi and see if it completes.

Sounds like a cold bug tbh

 

[Jleo]

My Amd64 and Dfi mobo are a real heart ache with Super pi, cpu voltage has to be just right otherwise I get your scenario over and over again.

 

[t31os]

Had the same thing when i didnt give me Venice core enough volts. SuperPI is fine now i've upped them.

Don't know if that helps.