NEW config - BSOD 0x124 and possible hardware problem ...

[patrack]

Hello all,

I just made a new config and having weird issues.
I have a Shuttle SH67H7 with a Core i7 2600k, 4x4Gb Corsair XMS, OCZ Vertex3, MSI GXT580 twinfrozr II. The basic PSU has been replaced by the optionnal 500W from shuttle.

This is a fresh windows 7 x64 Pro install.

After installation, windows patches and drivers installation, everything seems to work correctly.

I can quite easily reproduce the BSOD by simply running the "Windows Experience Index" check.
Yesterday, i tried to remove the ram and tests them 2 by 2. I managed to run the assesment some times without crashing the system, but it appeared again quite fast.

I did manage to run Furmark a bit to see if that was link to a high temperature, but it goes around 65°C with no problem, same for CPU with Prime95. I did not run these tests more than 10- 15min because the windows experience index test crashes the system faster, with temperatures not higher.

Here are some questions I have for you :
- Do you agree that this kind of configuration is OK ? ( GPU / box / ram /psu ...). I saw that Crazy Fool, a member here has nearly the same config..
- I just realized that I directly installed nvidia drivers and not MSI ones, does that change anything ?
- what should i do now to identify where that problem should come from ?


Thanks in advance for your help, I hope i'll find something

[GoNz0]

oh the joy of unknown hardware errors.

make sure the ram timings match the spec on the stick, corsair tends to throw a tizzy if its left on default timings.

[Phage]

My first thought was thermal as well.
At what point in the WEI does it crash ? Could it be a bad HDD ?
Might be a case of swapping out RAM, GPU etc one at a time.

[shaithis]

At what point in the WEI does it crash ?

Indeed. Knowing this (and doing multiple runs to ensure it is/is not crashing at the same point) is crucial.

[Millennium]

I think WEI crashing is pretty hard to pin down. Certainly I would be trying to get it to crash in other more meaningful ways... intel burn test? memtest86 for 12+ hours? this would really help us out.

I am going to take a punt now though and suggest some sort of hardware problem. If you have another GFX card swap it out and try that maybe (big maybe) your PSU is marginal with those components.

[patrack]

Thank you all for these replies.
I'll do some more tests tonight and this week-end.
As i do have another RAM set and a spare GFX EN210, i'll also try to swap ..
Same for memtest, not done yet.

[Ginger-Menace]

did you update the firmware on the vertex 3? i know we were having problems with our vertex 3 at work, and ocz claim to have fixed it with the recent update

[patrack]

To confirm, the SSD firmware is up to date..
I've done some tests, not a lot but they seems clear :
A memtest crashes quite fast with any memory module AND the GTX580. Doing the memtest with the onboard card runs perfectly for some hours. As soon as i try with the GTX, it crashes. --> Unexpected Interrupt - Halting , after 3-4 minutes. I can send printscreens if you want

The graphic card has just some days, i can easily return it. I just hope that I won't have the same problem because of my config, do you think I have to change the card ? 570 or less ...

[GoNz0]

did you manually set the ram timings or not ?

[Millennium]

Firstly I think you will need to get a less power hungry GPU as a couple sites i read say you need 600-650 watt PSU for it. Not sure this is 100% accurate just what the sites say.

However, it is strange that it is causing memtest crashes even then - when memtest is running it shouldn't be using all max watts since it isn't in a 3d mode.... (someone correct me if I'm wrong - it's in command line after all). Decide what GPU you would be happy with and then check here first to see if it would fit your PSU specs.

By the way I hope you don't plan to overclock that CPU much too. You didn't mention, are you overclocking anything now?

[kalniel]

0x124 is usually VTT/QPI voltage or vCore voltage. If nothing is overclocked it could just be your power supply/motherboard combination just aren't giving the CPU enough umph. If you increase the vCore and VTT/QPI voltages a little you should prevent the problem.

[CAT-THE-FIFTH]

Make sure you upgrade to the latest BIOS as there were issues with Turbo Boost with earlier revisions.

Also,Shuttle motherboards are very picky about the RAM they will use.

[GoNz0]

0x124 is usually VTT/QPI voltage or vCore voltage. If nothing is overclocked it could just be your power supply/motherboard combination just aren't giving the CPU enough umph. If you increase the vCore and VTT/QPI voltages a little you should prevent the problem.

very rare if your not overclocking and the rigs set to auto as it tends to overvolt anyway.

[patrack]

@GoNz0: i did not set the timings manually, i don't know how to do it and prefer not to play with it.
Same for overclocking, i've not touched anything. Not for me ... By the way, the gtx580 is the overclocked edition.
The BIOS is already up to date.
Concerning the PSU, i thought it would fit as per shuttle "power calculator". It says less than 400w for the config i have. Apparently, that is not really true ...

For the RAM, i'm currently testing with a kingston set from shuttle HCL .

I just tried also to add some voltage, as proposed by kalniel, and this time the windows experience did complete. I'll do some more tests. Sure that a COD MW3 will be a good one .

Thanks again for the tips !

UP --> Nearly one hour of game@2560*1600 full options with no problems !!! voltage trick helped !?

[Jonj1611]

I had that exact same error from a Q6600 that had been a stable overclock for 2 years, turned out the voltage needed to be upped as already suggested, its been stable ever since

Try running Prime95 for a few hours mate

[GoNz0]

prime95 is outdated, intel burn test is the better option.