CoD W@W is crashing in Win7

[Zak33]

I've got a problem... CoD W@W was rock solid in XP and Vista 64 but I'm getting a few crashes in Win7.

Anyone know what this is:

Problem signature:
Problem Event Name: APPCRASH
Application Name: CoDWaWmp.exe
Application Version: 1.7.0.0
Application Timestamp: 4aea1f42
Fault Module Name: CoDWaWmp.exe
Fault Module Version: 1.7.0.0
Fault Module Timestamp: 4aea1f42
Exception Code: c0000005
Exception Offset: 001e3641
OS Version: 6.1.7600.2.0.0.256.48
Locale ID: 2057
Additional Information 1: 454d
Additional Information 2: 454d32af79114f9d8278258870b48803
Additional Information 3: d564
Additional Information 4: d564b07e1478d8e937292aa2c31d79ad

Done it twice tonight in half an hour. First time for a week I've had a spare hour and I can't play

[kalniel]

General protection fault. Could be any number of things, but very likely to be a driver that's just not as compatible with windows 7 as the others, or you've simply not updated. Check printer/NIC/sound card etc. for updates.

[watforddude]

Most crashes I've had have been due to the sound driver, is yours realtek?

with COD W@W that is

[GoNz0]

did u update directx ?

[nibbler]

Pure hax if you ask me

[Zak33]

I agree about the sound card.. i've got a top line card but with crap drivers.

Razer Barracuda AC1. with Vista 64 bit beta drivers cos they never quite sorted them... was fine in Vista64.

[Zak33]

did u update directx ?

I did, yeah, but it's a good thought.

[jim]

Why don't you enable the onboard sound, disable the discrete card in device manager, and just test it out for a while?

Or have you already done that?

No idea why it would be the sound card in particular, but crap drivers does make it somewhat more likely.

[CAT-THE-FIFTH]

It seems more like CoD Win7 at War.

[Zak33]

have tried onboard sound... didn't solve it.

but.. I think it's my overclock.

My QX6700 2.67 at 3.2 is rock solid in CPU burn in on al 4 cores for 4 hours...

but Iknocked it down to 2.9 and it's not crashed at all... think it was a CPU overheat drop out issue... played for 3 hours tonight.. only crashed out when I upped it to 3.2 again.

Don't think this cpu likes higher clocks... got a superb heatsink and at defauklt volts it's rock at 2.9... but at 3.2 it's flaky with extra .05 and .01 and .15 volts... tried em all.

[Tonka777]

4 Hour stress test Zak? Gotta be 24+ to be stable for me.

[Zak33]

4 Hour stress test Zak? Gotta be 24+ to be stable for me.

hmm.. I understand people's love of 24 testing, but if the cpu reaches a lower heat in gaming in 15 minutes, and crashes out to desktop...than it does under 3 hours of cpu burn in on all 4 cores, where is goes higher temp....

well, that's what confused me.

[kalniel]

hmm.. I understand people's love of 24 testing, but if the cpu reaches a lower heat in gaming in 15 minutes, and crashes out to desktop...than it does under 3 hours of cpu burn in on all 4 cores, where is goes higher temp....

well, that's what confused me.

Which bit of the cores is being stresses for each respective test? Cache utilisation the same? etc. Just because repeatedly running the int pipeline with a small amount of cache usage loads up the temperature it doesn't mean that some other part of the chip won't be unstable at that frequency.

[watforddude]

At one point if I overclocked my GFX card even 1Mhz past its already overclock it would crash/go crazy etc! just a thought try setting it at 3.1 and see if it still crashes that way you haven't lost to much? just a thought

[scaryjim]

Could the memory be the problem rather than the CPU? Trying running memtest at 3.2GHz and 2.9GHz? Or did you just reduce your overclock by dropping the multi one notch?

[Tonka777]

hmm.. I understand people's love of 24 testing, but if the cpu reaches a lower heat in gaming in 15 minutes, and crashes out to desktop...than it does under 3 hours of cpu burn in on all 4 cores, where is goes higher temp....

well, that's what confused me.

Its not a love of 24+ hour testing, its a necessity for me. Its also not about how hot i can get the CPU, the heat is a just a by-product of the stress test, so IMO it doesn't matter which app you use as long as the CPU stays within thermal limits. The point of running the test for very long periods is to ensure all cores consistently calculate the sums correctly and that all of the memory is used in this process. With enthusiast systems having more and more RAM these days, it goes without saying that the stress testing should take longer, just to ensure all of the memory is tested. I don't know, but can you be sure that all of your RAM was tested in those 3 hours?

Not trying to be condescending btw, just trying to explain why i think it necessary to test for long periods. If your interested, i use Prime95 alongside as many instances of HCI memtest is needed to use all of the RAM. I let prime run constantly, and i try to make sure HCI says it has tested 100% of the its allocated RAM 3 or 4 times over.

I use linpack for quick tests during the overclock process, as it seems to find the errors quicker, but it is not fail safe for me. I ran 50 runs of linpack last time i OC'd and passed, but then failed Prime95 18 hours in. I cant see the point of running the CPU that hot for 50 runs, to find out it is not stable whilst gaming afterwards.