Prime95 25.11 & 27.7 versions and IBT 2.54

Associate
Joined
19 Aug 2011
Posts
149
I have my System now what I think is stable 3570K @4.50GHZ 1.256v with a + Offset of 0.035v Temps 71c MAX!

However! Prime 95 is rather weird between these 2 versions, 25.11 reports Core Voltage on CPU-Z as 1.256v and Prime95 27.7 reports it as being 1.296v :rolleyes: both these decide to stop after 2 hours on Worker 2 so I decided to give IBT a shot, With Success! 30 passes on Standard, and 10 Passes on High! all Pass flawlessly, yet Very High doesn't kick it, it will error after 2 passes.

So........ having said that, would anyone else consider this to be stable under the circumstances?

Another thing, if a persons PC passes and is is stable under a Prime95 test, and fails with a new version of Prime95, was it really stable in the first place? This could get rather Paranoiac, if Prime keeps releasing versions that generate fails, that were thought to be stable by the previous version. With that said, I'll reject Prime95 and declare it unstable to test CPU's with for real world results.
 
Associate
Joined
6 Apr 2006
Posts
1,383
It seems likes its not quite stable, I'd use it for awhile and see if in your normal use any problems crop up, I find I can complete IBT and prime for many hours but crash in BF3 and have to up the voltage.

Basically don't rely on stress testing.
 
Soldato
Joined
24 Mar 2008
Posts
4,654
Location
High Wycombe
Agreed, passing those stress tests just means its stable enough to run them though they can give a very good indicator of long term stability.

I wouldn't say your PC is stable though and will need to up the volts. Although you may not see instability initially you will be writing errors to the hard drive and in time start getting program crashes, CTD's and BSOD's
 
Soldato
Joined
3 Jan 2006
Posts
5,009
If I remember correctly prime95 27.7 is the latest version with AVX support so it's more demanding on latest cpus that support this instruction, will probably explain the voltage difference as well. What load line calibration setting are you using?
 
Associate
OP
Joined
19 Aug 2011
Posts
149
If I remember correctly prime95 27.7 is the latest version with AVX support so it's more demanding on latest cpus that support this instruction, will probably explain the voltage difference as well. What load line calibration setting are you using?

I set manual adjustment, so doing this changed Optimized Power Phaze to ultra Fast, So Ultra High, and Ultra Fast in LLC. If I can get the volts right, I may have to change to a better cooler, maybe a corsair hydro 80, I have the Noctua Performance cooler at the moment, and under Prime kicks out 86c + when Prime's running the 8k test, for anything above 1.256v
256
 
Last edited:
Top Bottom