Windows XP SP2, Intel QX6700, 8800GTX, 2gb ram, BSOD minidump.

[Shigawire]

I've suddenly been experiencing some nasty BSODs, and I was wondering if someone could please take a look at the minidumps.

Computer is a
Windows XPSP2, Intel QX6700, 8800GTX, 2gb ram, on a Asus Striker Extreme mobo.
I also have various USB devices.

I see on my Asus PC PROBE II that the voltage for "CPU VTT" is going into the red areas fairly often.
Whereas nominally it'd be between 1.25 and 1.35, mine goes from 1.25 down to 1.09 and then goes up again.

2 minidumps are attached.

Thanks in advance for any help that you kind sirs may provide.
Techspot is the best place for true expert advice.

 

[Route44]

Okay, both minidumps are too general in citing the faulting driver that can give us a definitive answer. But both also point out that a driver/process is attempting to access an IRQ address that it has no business doing.

The thing that catches my eye however is your statement of voltage irregularity and RAM. Two posters this week have had issues with their RAM voltage and BSODs. Both overclocked and both had to back off of their RAM timings.

One process of interest that did show up in one of your minidumps was boincmgr.exe which is a driver for BOINC Manager for Windows belonging to the software BOINC Manager or CPDNBBC core client or boincmgr.exe by Space Sciences Laboratory, U.C. Berkeley.

I take it you do folding?

My best first guess is that if you stabilize your RAM voltage you'll stabilize the system. The fluntuations concern me. Bad PSU? ASUS has been known to be very picky about RAM as well.

Have you run MemTest? What is your memory brand?

 

[Shigawire]

Yes I do a bunch of BOINC-stuff
I say, why let all that juice go to waste when I'm not using it?

WorldCommunityGrid mostly. Help Conquer Cancer, as well as Proteome Folding stuff.

First off, I don't overclock my PC. The 65nm design is already too bulky for the boxed HSF. The QX6700 was indeed way too hot when all 4 cores are running simultaneously, even for the boxed fan. I've got a liquid cooling setup.

Occasionally the temperature goes up to 65 celsius (the stated limit of the CPU). It has even gone higher than 65, reaching 66 and 67 celsius. Do you think the CPU, or other components, could have taken damage at this level?

PSU is Seasonic M12 700W.
RAM is Corsair Dominator TWIN2X6400C4D 2048MB.

So.. this "CPU VTT" is ram voltage? It's fluctuating from 1.07 up to the 1.40s.
Isn't there a way to clamp it down?

I have not yet run memtest in recent time. But I have run memtest before on this same system and ram.

edit.
a recent observation: some slight static from the speakers.
could this be related to the fluctuations in voltage?

 

[Route44]

Seasonic psu's are known for their quality and Corsair makes some fine RAM.

That heat is way high! 60-70 degrees Celcius = 140-158 degrees farenheit!

You need to do some serious cooling A.S.A.P. I am suprised you haven't had shut-downs. High Heat is know to bring instablity.

What are you doing to cool the system now? How is the air flow.? Are ALL fans working?

 

[Shigawire]

Seasonic psu's are known for their quality and Corsair makes some fine RAM.

Yes I think so. I bought the Seasonic for its good review in www.SilentPCreview.com
And the ram looked fine.

That heat is way high! 60-70 degrees Celcius = 140-158 degrees farenheit!

Yes.

You need to do some serious cooling A.S.A.P. I am suprised you haven't had shut-downs. High Heat is know to bring instablity.

Yep I've opened up the liquid cooling system, and taken the computer out of the desk. There were some gas bubbles in the hoses. Idle temp used to be way high, like 47 celsius. Now after the degassing the idle temp is 34 celsius, the way I remember it.

What are you doing to cool the system now? How is the air flow.? Are ALL fans working?

I have no fans in the puter. I have a fan outside, a regular desk fan pointing to the Zalman Reserator 2. You see, the Reserator 2 is not strong enough to cool the required amount, so the desk fan cools it a lot better.

In the computer I have only the PSU fan, which is working fine.

I think I need to upgrade the liquid cooling system from Zalman Reserator 2 to something like the Zalman Reserator XT.

Some things I did before degassing:

I plucked out both the RAM chips, unplugged the power cords, blew some air with my can of pressurized oxygen in case some dust was wreaking havoc in there.
Then after the degassing I reseated the ram and powerplugs. But only 1 ram to keep it simple.
Now the CPU VTT isn't as volatile as it was before. Now it goes from 1.15 to 1.25, and it usually holds well on 1.25.

Of course this could also be relating to the temperature being lower, or there being only 1 ram chip in there, or some dust causing the problem.

I'm going to put the other ram chip in now.

 

[Route44]

Ah, yes, dust. Talk about a "natural" blanket for keeping in heat...

I would be curious to see if you put that other stick of RAM in now to see if the voltage regulation still remains constant.

It sounds like you've regained stability through proper cooling. 34 Celcius is a good safe temperature. Keep us up to date; I know I would love to read how it works out.

 

[Shigawire]

Yes sir. I don't like dust very much.

The other ram is in as well now.
CPU VTT voltage is not as volatile as before, but it's between 1.15 and 1.25. Mostly on 1.15

I also installed a Noctua NF-R8 case fan to help the airflow inside the computer.
Seems to be a good idea. The temperatures are lower on both the MB and CPU.

However, I'm doing some extensive BOINC tests now, and I'm not sure the CPU is working the way it should. I don't remember how the CPU Usage went last time I ran BOINC. I've set WorldCommunityGrid to use 100% of all 4 cores.

I see what I consider some really weird behavior now. The CPU Usage goes up to 100% , and then down to 25% then up to 100% then down to 25%.. this is on all four cores. Each of the 4 cores are running different BOINC tasks. Two are running Help Conquer Cancer, the other 2 are running Proteome Folding Phase II.

I have attached a picture of the cpu fluctuations. I just don't remember if this was how the folding actually affected the CPU.. so I can't tell whether it's an anomaly or not. My gut tells me it is.

 

[Shigawire]

Ok, I suspended the 4 BOINC operations and ran BOINC's own "CPU benchmarks" - and there it surely used 100% all the time while running the tests.. with no flux at all. So the up-down flux may just be the way the Proteome and Cancer calculations are handled.

 

[Route44]

This is good to hear. If stability has returned through correct cooling then it looks like the issue has been resolved. Thanks for telling us the results and good luck!

 

[Shigawire]

Yep. The cooling is a lot better. The max temperature now is 50-58 celsius. Mostly 54-56. I do have some issues with my Windows XP though, so I will have to reinstall it regardless. You see, my USB ports are not functioning correctly. Windows XP refuses to recognize new devices connected to USB. I have a WACOM INTUOS3 A4 drawing tablet, which I paint with. And installing it is impossible now.
First the found new hardware wizard comes up. Then after searching for the driver, It just says
""Can not install hardware, not recognized"

The same happens if I insert my usb keyboard into the "wrong" usb hole.

I believe this issue is unrelated, and was before I ever had any heat issues.

 

[Route44]

It could be a BIOS/chipset thing as well.

 

[Shigawire]

Well it wasn't like this before, on the same computer. So it's software-related most likely. The operating system is my guess.