• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Prime95 Error

bob4432

Lifer
in the torture test i can run the "Small FFTs" & "Blend" test forever, but the "In-place large FFTs" for only about 5mins. i have run memtest for 10 runs with no memory errors, what is the in-place large ffts error telling me?

thanks
 
I assume you are overclocking.....
If you are, you have your PC set @ a faster clock than your CPU can handle without bumping up it's voltage. You could be having heat issues as well. Prime95 stresses your CPU bringing temps near max. (about 98% utilization)

There are a few different probs you could be having, but there is no way to diagnose without more information.
 
It's most likely that your cache is overheating. But it's somewhat strange that the Blend test didn't do anything. To get better answers, you might try the Prime95 Q&A forum.
 
actually i am not o/c, i just got this board back from an rma. i had bought it really cheap and it turned out to be doa, but since i got such a good deal and it was still under warranty, i rmaed it to msi and just got it back last week.

i wanted to get a baseline before i o/ced it. i am using a decent copper hs with a 92mm fan and have cpu temps ~35C idle and 52C under load. i just pulled out another i865pe based board that i was running and the temps are inline with what the other board was reporting, so i am ok on the temp issue.

one thing i did notice was in the bios i just upped the cpu voltage little bit (don't know the exact number, but in the bios i think it went from 1.525 to 1.535 or something really close) and my error went from "received a value over 5 and was expecting a value less than 4..." to a value that was over 4 but in the 4s, like 4.xxxxxxx, so the cpu voltage did do something to it in that regard.

should i just up the voltage on the cpu? do yo know how high i can go with a 2.8c?
 
maybe i will just put my old m/b in another spare case i have and swap out my cpu/ram and see if the board is the problem......
 
Well, You could try upping the voltage a bit....
Some boards tend to undervolt somewhat.
Keep an eye on temps when you are adjusting cpu voltage.
As long as you take baby steps you should be fine. The Athlon XP in my system has been overvolted .125v for over 2 years now and has never had any problems.

I would definately try isolating components though.
@ stock settings there shouldn't be any reasons to have to overvolt.

It also could be that your PSU isn't providing stable power.
You might check how consistent your voltages are when running Prime95.
 
stock setup, according to mbm5 cpu is between 1.49 and 1.5V, and everything thing else stays with .05V, except the 12V line which states 12.15-12.28V...
 
Originally posted by: bob4432
stock setup, according to mbm5 cpu is between 1.49 and 1.5V, and everything thing else stays with .05V, except the 12V line which states 12.15-12.28V...

Are those your readings while runnig Prime95?
If that all the variance you get when stressing your system (Prime95) your PS is fine.

Are you using Prime95 V23? This is the latest "official release".
Beta versions tend to crash.

I would probably try upping your vcore a little more and keep an eye on temps and fluctuations on your vcore and vdimm voltages. You shouldn't be crashing is these are stable and memtest runs error free.

Trust me, you'll know when you hit a voltage wall. Temps will start to rise sharply.
 
thanks for the reply, yes that is with prime95 running. i have been using i think the newest version, i will d/l again to verify, but i think mine is 23.8...even when i did adj voltages in the bios, mbm reported basically the same voltages as this, and there have been no sudden spikes in temps. i will leave memtest on all night tonight again to verify...
 
i messed around some more and have run f@h for hours today and all thing work ok, except prime95....if i o/c the cpu and up voltage, i get the error faster...
 
Originally posted by: bob4432
i messed around some more and have run f@h for hours today and all thing work ok, except prime95....if i o/c the cpu and up voltage, i get the error faster...

Well, I definately wouldn't oc your processor until you start getting stability using Prime95. Matter of fact Folding will probably abort wu's occassionally until you get stable.

You might want to take this to the overclock forum. You will probably get more feedback from people running similar rigs.

On my main box, when I initially set it up, I was using onboard graphics and had to up my CPU, chipset, and vdimm voltages to get stability. (oc'd of course) Later when I installed a 980 Pro, I was able to reduce my chipset voltage, but still have a .125 vcore and .1 vdimm bump.

My only other suggestion would be to update your bios and drivers to latest versions if you haven't already.
 
everything is up-to-date, i will probably call msi on tues and see what they say because i can swap the components into my other board and be fine...but is just a pain in the a$$...thanks for the info 😉
 
just for sh!ts and giggles i put win2k on it to see if my install was ok. i turned on the small ffts and i stopped it after an hour, everything ok, i did the large ffts and sometimes it will run for an hour and i will stop it or it will run for 20mins and error out....since i just changed m/bs, do you think that could be the issue?
 
just a quick update - looks like a bad stick of ram...even though it ram memtest the first time fine. i have now tried this certain stick in my i865pe board and also a nf2 board and get errors out the ying-yang in memtest. running one stick in my i865pe board and running the same prime95 test and no problems. will let it run all night and see what happens...
 
Originally posted by: Rastus
Sometimes it helps to move memory around in the slots.


i did do this with the one stick and it has been able to do the prime95 torture test for 7hours+ at which time i stopped it. i will call ocz today to get an rma for this other stick since in ~3hrs in memtest i get ~700,000 errors 🙁

what could cause one stick to go bad and not the other?
 
Back
Top