Random pc shutdown

[ortega10]

I have bulit a pc that i keep at work. When i am at work the pc doesnt have any issues and does not shut off at all. Whenever i leave work and come back i find that my pc has shut down abruptly. I know it wasnt shut down manually because i get the black safe mode screen when i start it up. It usually starts back up with no issues.

I am going to assume that this is due to some sort of overheating issue. I have a watercooling system installed tha goes, pump, scp, radiator, to reservoir. i have a flow meter so i can see that the water is flowing well. i noticed a small kink in the tuing this morning, which leads me to think that after a while the flow of the water may be slowing down and causing the cpu to overheat and subsequently shut down.

I was wondering if there is any way that i can be sure that this is what is causing the problem. I was thinking through the evenvwr.msc or maybe if i keep some sort of program running that could record some sot of log of what happened.

Keep in mind this pc needs to be kept onat all times because it suns te companies dvr s well as a couple of monitors showing traffic cams.

Any suggestion are greatly appreciated.

Specs:


corsair 1000

2 - galaxy gtx 460 in sli

core i7 980x

6 gb ddr3 ram

q-see 8 channel dvr (model # qsdt8pcrc)

2 ssd in raid 0 - intel 80 gb and 60 gb ocz vertex - in 4 ssd hotswap

1 lg blu-ray rewriteable drive

 

[HPC user]

I have bulit a pc that i keep at work. When i am at work the pc doesnt have any issues and does not shut off at all. Whenever i leave work and come back i find that my pc has shut down abruptly. I know it wasnt shut down manually because i get the black safe mode screen when i start it up. It usually starts back up with no issues.

I am going to assume that this is due to some sort of overheating issue. I have a watercooling system installed tha goes, pump, scp, radiator, to reservoir. i have a flow meter so i can see that the water is flowing well. i noticed a small kink in the tuing this morning, which leads me to think that after a while the flow of the water may be slowing down and causing the cpu to overheat and subsequently shut down.

I was wondering if there is any way that i can be sure that this is what is causing the problem. I was thinking through the evenvwr.msc or maybe if i keep some sort of program running that could record some sot of log of what happened.

Keep in mind this pc needs to be kept onat all times because it suns te companies dvr s well as a couple of monitors showing traffic cams.

Any suggestion are greatly appreciated.

Specs:


corsair 1000

2 - galaxy gtx 460 in sli

core i7 980x

6 gb ddr3 ram

q-see 8 channel dvr (model # qsdt8pcrc)

2 ssd in raid 0 - intel 80 gb and 60 gb ocz vertex - in 4 ssd hotswap

1 lg blu-ray rewriteable drive

Go into the BIOS. Look under the "CPU settings" tab (or something similar; depends on your BIOS). Disable the "C-states" flag. Your machine should then stop crashing. The C-state was developed by Intel to minimize the thermal load of the CPU whenever possible. Only, it makes the CPU unstable (leading to a kernel panic due to de-sync in worst case), see here:
http://download.intel.com/design/intarch/papers/323671.pdf

Irreproducible crashing of the system is the result. I disabled the C-state flag in my BIOS (Clevo 900F, i7-990X) and had a superstable system which never crashed ever since. Got a SIM job running on all six cores for seven weeks now -- all fine 🙂

Hope I could help.

 

[ortega10]

Go into the BIOS. Look under the "CPU settings" tab (or something similar; depends on your BIOS). Disable the "C-states" flag. Your machine should then stop crashing. The C-state was developed by Intel to minimize the thermal load of the CPU whenever possible. Only, it makes the CPU unstable (leading to a kernel panic due to de-sync in worst case), see here:
http://download.intel.com/design/intarch/papers/323671.pdf

Irreproducible crashing of the system is the result. I disabled the C-state flag in my BIOS (Clevo 900F, i7-990X) and had a superstable system which never crashed ever since. Got a SIM job running on all six cores for seven weeks now -- all fine 🙂

Hope I could help.

Thank you for the help I will definately try that. The pc still has random reboots so maybe this will help.

 

[HPC user]

Thanks for the message. Keep me posted how it goes!

BTW, in its white paper, Intel calls the CPU-desync caused by C-state switching "indeterminism" ... pretty euphemistic description for a random CPU crash.