• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

PC will instantly fail P95 at random times.

B

Bop

2[H]4U
2FA
Joined
Oct 1, 2003
Messages
3,426
I've noticed that my PC can run P95 for 12+ hours straight but sometimes it will instantly fail. If I reboot it will work fine.

Voltages seem to be the same during reboots so I'm a little confused as to why this is happening. Fiddling actively with the voltages in E-LEET don't seem to make any difference whatsoever to stability.

This is rather annoying but my apps seem to be stable. Could this be a glitch with P95?
 
I've seen it do that before on mine and a buddie's of mine. Are you just running SmallFFTs?
 
Well in that case Blend is going to hit the CPU and the RAM

Try running SmallFFTs for CPU and LargeFFTs for RAM
whichever fails you'll know which is causing the instability

but honestly, it may very well just be P95 being stupid
 
I ran memtest before and there were no errors after 10 hours. The next time I have this issue I'll definitely try SmallFFTs, that's a good idea. I wish I knew how to reproduce it. Just rebooted and its rock stable again.
 
yea P95 pisses me off like that a lot .... it will run solid for 8-10 hours then the next day it fails after 2 minutes lol
 
SmallFFTs = ok
LargeFFTs = Failed after 9 minutes

Blend = Instant fail.... motherboard issue? Rebooting will fix it :/ I'm heading to radioshack to get a multimeter to check the voltages.


EDIT: I noticed that your board and memory is from EVGA and G.Skill. Does your friend have the same setup?
 
Last edited:
Psu instability?

I had some issues with p95 failing when I had some of my voltages set to auto. I always set my cpu, and ram at their voltages they like, but stuff like the fsb and other others I left auto.

FSB likes to be set, and really I am thinking the others as well. so stopped using auto mode
 
Also, when P95 fails, please tell us what the error is.

Usually, if its a error rounding like came back with a .04 should of been a .05 that is almost always a cpu error.

And if it says the sum of 30434050304500.034004 should be the sum of some other number. (I call it the sum of the sums) then its usually a ram error.

Those are the 2 I know. from using prime alot back in the day.
 
I ran a multimeter on the motherboard testing points and the voltages are normal. VTT was undervolted by .1V but increasing those volts did nothing. I had to boost vdimm as it was falling to 1.63 under load but that did not do anything to stability.

I'd test my PSU next but I'm a novice when it comes to that and I don't know where to poke around.

Also lowering my OC 450mhz with E-LEET in windows does nothing. It still instantly fails despite everything being downclocked. Which makes no sense to me. My RAM timings are tight but I can confirm the problem can still occur with CAS 8 timings. It passed memtest overnight at 6.

Linx fails in seconds too. It took two reboots to get it back to normal.

Could the motherboard be the problem or is it maybe the CPU's memory controller? The fact that rebooting fixes it makes me think this is a motherboard issue. I had an MSI that did something like this with an old AMD CPU.

UPDATE: I went to G.Skill's website and they apparently changed the memory specifications for my kit from 1N to 2N. My command rate is 1N so I'll see if 2N affects anything. I guess they retroactively noticed a problem with the memory and/or the memory controller? I hope this works...
 
Last edited:
It didn't, unfortunately.

It seems to happen in the morning. Rebooting in the morning doesn't fix it but after I go into the BIOS and change a setting like Uncore Frequency and back it seems to work ok.

I still don't know where the problem lies but it seems like a motherboard problem. I may RMA it and sell it anyways with the new SATA 6/ USB 3.0 boards out now.
 
bump, I tried it in a new board and it still occurs. This really seems to be a first boot in the morning thing. It makes no sense unless my PSU needs to "warm up."
 
Bop said:
bump, I tried it in a new board and it still occurs. This really seems to be a first boot in the morning thing. It makes no sense unless my PSU needs to "warm up."
Click to expand...

Well, sounds like a MCH (which is integrated into the cpu) instability. I had similar things happen when going from 4GB to 8GB of ram with my q9650. Four occupied dimms and a quad is very stressful to a p45.

I'm not too familiar with i7 so I can't really help you out.
 
BababooeyHTJ said:
Well, sounds like a MCH (which is integrated into the cpu) instability. I had similar things happen when going from 4GB to 8GB of ram with my q9650. Four occupied dimms and a quad is very stressful to a p45.

I'm not too familiar with i7 so I can't really help you out.
Click to expand...

Possibly, I'm going to try my luck with a D0 i7 920. If it works then I'll RMA this CPU and sell it. I tried another PSU and it still occurs so it's not the power.

The last memory I used it would pass Prime95 at DDR3 1600 but would freeze 3d apps. Maybe it just can't handle 1600 speeds.... but with my EVGA board I could lower the BCLCK in Windows and that didn't do anything to help stability. Maybe I just have a dying CPU, which is strange b/c I never over-volted anything. Shit luck I guess.
 
Bop said:
Possibly, I'm going to try my luck with a D0 i7 920. If it works then I'll RMA this CPU and sell it. I tried another PSU and it still occurs so it's not the power.

The last memory I used it would pass Prime95 at DDR3 1600 but would freeze 3d apps. Maybe it just can't handle 1600 speeds.... but with my EVGA board I could lower the BCLCK in Windows and that didn't do anything to help stability. Maybe I just have a dying CPU, which is strange b/c I never over-volted anything. Shit luck I guess.
Click to expand...

Is this happening at stock speeds? I would try pulling back on the trfc, and playing with the mch gtl but like I said I'm not at all familiar with i7.
 
Wow, I figured out what caused it but I can't explain it and I don't know what kind of problem it is.

It is caused by TIME. From 8AM-9AM it will instantly fail. If I move the PC clock to 11AM it will work! If I move it back it will fail again!

I've never seen anything like this... what the hell!?

Should I replace the CPU?

EDIT: Spoke too soon, happened again at a later time. This occurs every morning though. Linx passes now but Prime fails. If I start a test sometimes it will run, but if I stop it and run it again it will fail. If the test doesn't fail in the first few seconds it will go for hours without problems.
 
Last edited:
Bop said:
Wow, I figured out what caused it but I can't explain it and I don't know what kind of problem it is.

It is caused by TIME. From 8AM-9AM it will instantly fail. If I move the PC clock to 11AM it will work! If I move it back it will fail again!

I've never seen anything like this... what the hell!?

Should I replace the CPU?

EDIT: Spoke too soon, happened again at a later time. This occurs every morning though. Linx passes now but Prime fails. If I start a test sometimes it will run, but if I stop it and run it again it will fail. If the test doesn't fail in the first few seconds it will go for hours without problems.
Click to expand...

Can you try this out for me? This test is very, very stressful on the MCH. Lets put it like this I was passing long runs of Prime small fft, blend, linpack, and memtest 3.8 and failed this in minutes. I hear that this is acctually more stressful on i7 than on a P45, too. :D

Start Prime95 x64, then start a torture test and select custom, under min FFT enter 128, then under max fft enter 1024, then enter all of your available ram where it says memory to use. Let that run for a couple of hours.

If prime freezes up that means that you are unstable, it could also fail on a core as usual, or even make your rig bsod. If you pass this it rules out your MCH.
 
BababooeyHTJ said:
Can you try this out for me? This test is very, very stressful on the MCH. Lets put it like this I was passing long runs of Prime small fft, blend, linpack, and memtest 3.8 and failed this in minutes. I hear that this is acctually more stressful on i7 than on a P45, too. :D

Start Prime95 x64, then start a torture test and select custom, under min FFT enter 128, then under max fft enter 1024, then enter all of your available ram where it says memory to use. Let that run for a couple of hours.

If prime freezes up that means that you are unstable, it could also fail on a core as usual, or even make your rig bsod. If you pass this it rules out your MCH.
Click to expand...

It didn't fail. But this was after the instant failures stopped occurring for the day. I turned my PC on past 10AM and now it seems it isn't happening. Something about turning it on at a certain time of the day... either way it is either the CPU or the RAM at this point.

EDIT: I tried a D0 i7 920 and the same problem occurs. So... by the process of deduction the sole culprit is the RAM. I've had problems with G.Skill in the past too.
 
Last edited:
You must log in or register to reply here.
Back
Top