3770k failing again

[masterg]

so my 3770k system is dying again. first it would heat up and bsod due to GPUs getting hot. or it would freeze and lockup the desktop. i pulled them out, cleaned them and reseated them, cleaned out H100 rad too. i will get an aftermarket cooler on when i get time. started it up and still seizing when the gpus run Milkyway@home. I disabled MW@H and i wake up this morning to a BSOD thats stuck: 0x0000009C MACHINE_CHECK_EXCEPTION

have no idea why this keeps freezing and dying. this is how it started the last time my ssd failed tho

[retro77]

Do you think its some faulty hardware? How does it run in default BIOS settings?

[andressergio]

Do you think its some faulty hardware? How does it run in default BIOS settings?

agree with the mate try stock, you pulled the IHS of the 3770K ? if not do so willl drop 10C certified

freeze could be many things first MEM, run a memtest and let it pass 400% 4 instances at 1500RAM if you are runing 8GB

but first of all go stock and play watch movies give it a heavy use ok ?

then report back

also can be BIOS corrupted u can try to reflash again

cheers and best of lucks
Sergio

[masterg]

cpu temps are mid 60s running a push pull H100 on it. GPUs were hitting 80c and thats what was causing it to crash earlier

[andressergio]

cpu temps are mid 60s running a push pull H100 on it. GPUs were hitting 80c and thats what was causing it to crash earlier

take the IHS replace tim and reseat 10C less certified and more stable

you got all on the Gene 5 thread

or PM me

cheers !
Sergio

[masterg]

I really dont have the money to make sure i don't fuzk it up, so id rather not try.

[andressergio]

I really dont have the money to make sure i don't fuzk it up, so id rather not try.

its so easy that u cannot up anythin mate it goes off so fast you will not notice but as u wish u want high clocks on and on AIR ? then u must do it

[dengyong]

its so easy that u cannot up anythin mate it goes off so fast you will not notice but as u wish u want high clocks on and on AIR ? then u must do it

H100 is a corsair water cooler with a double rad.

[Sparky]

If he's running in the 60s temps shouldn't be the issue. A bit warmer, yes, but temps alone shouldn't be causing the BSODs.

[masterg]

I have a 5830 and a 5870 vapor-x in crossfire with the 5830 on top. right next to each other, 5830 is clean and its speeds have been halfed from 800/1000 to 400/500 and still pushing 80C. i turned off the gpu crunching and its fine for now, i think the 5830 needs a new heatsink, the fan could be dead for all i know, cant exactly see it when its all running

[andressergio]

H100 is a corsair water cooler with a double rad.

yep didnt read that thanks

[masterg]

event logger says its a power management issue on core 0. and parity issue

Performance power management features on processor x in group 0 are disabled due to firmware problem. Check with the computer manufacturer for updated firmware

[Conquistador SW]

I never heard of that power managment issue before.

I have some experience with the WHEA error parity issue though. I noticed that even though Linx/prime95 will never fail and so you think it is stable at a certain voltage, there will still be some WHEA errors until you raise the voltage further.
My 3770k was linx/prime95 stable at 4.7Ghz at 1.26v, but the WHEA errors only disappeared when I got to 1.31v.

Maybe raising the core voltage will help solve your problems.

[WhiteFireDragon]

GPUs were hitting 80c and thats what was causing it to crash earlier

How is 80C even considered hot? I ran my gtx295 for several months 24/7 at 105C doing gpu crunching for gpugrid, knowing that it's a suicide run but somehow the video is still perfectly fine lol. I also did bitcoin last year for a little bit with AMD cards, 80C is pretty normal. This is with 4 cards stacked on one mobo. I don't think that temp is what caused your crash.

[Conquistador SW]

Indeed, with my HD5870, the fan profile only started to really ramp up the fan when it hit 80C. It would only reach max rpm at 90C, so AMD apparently thought 90C should be an acceptable max temperature.

[masterg]

ok, maybe it wasnt gpu then. it just seemed like it was the most obvious answer. event logger says otherwise

[Sparky]

If it does this even at stock speeds I'd start suspecting motherboard or the CPU itself even. I've had a defective CPU in the past.

[masterg]

its running 99.5x44