New PC - BSOD due to having Blender open

Support Technical Support
1

I have been donating cpu power for renderfarm projects like renderfarm.fi and burp.renderfarming.net for some time now. I just recently bought a new 8-core, 32 Gb RAM pc for 3D editing/rendering/renderfarming, and when working with Blender it crashes the entire system at least twice a day (like every 6 hours). I am 1000 % sure it is not due to overheating and also pretty sure it is not due to the motherboard failing.
I have been running the pc for about 5 days now without ever opening up Blender (yep, it has been tough not working with Blender at all :P) and my pc has not crashed once!
I have no idea where to go with this problem as I’m having a hard time figuring how Blender can crash my pc, I will be asking around on different forums of what to do, but what do you guys think?

Cheers

2

Try testing your memory modules, this is usually the case, either on the mother board slots or the modules themselves. You can test them by running memtest86, I belive it is. There are cases where a program my try to address a bad area of a memory module. The other possibility is your video card.

3

If you are over clocking, try lowering the speed.
A true BSOD is Windows crashing due to hardware/driver problems. Blender is simply revealing the hardware problem. Which means you can use Blender to help determine when the hardware is stable.
What is the rating on your power supply? If you do not have enough power, that can cause crashes, especially when an application starts to make use of otherwise unused features, such as 3D hardware.

4

No one single program can crash hardware or OS. Obviuosly you have faulty hardware/drivers/OS, blender just use that faulty system call that no one other program use. Try cpuburn and memtest like programs, better that not require any OS and run from LiveCD, run them for days, see result. Most probably it is faulty RAM, PSU voltage fluctuations, overheat.

5

I will test my pc using memtest84+, getting back later with the results

I have an eVGa 3Gb standard-overclocked 660ti, I don’t know how to test if it is the problem.

6

I am not overclocking (apart from standard manufacturer OC) atm, will be overclocking later when the system is stable

I have a great 650W power supply from XFX, my pc uses up max 220W atm, will change when my 640 GT arrives though

7

I have been monitoring my pc through HWMonitor and AI Suite II for my ASUS motherboard, I could try the LiveCD tests too

8

I let my pc run through 2 passes with memtest, no errors found so far. I will have it run through the night, making sure, it’s not the RAM that is the problem

9

Memtest does not check for (thermal) stress, only for physical error.
I’d recommend to use Prim95 and kick your system in the groin with it.
Run all the tests, CPU burn in, RAM stress-test and so on.

I’d say 99% of BSOD from RAM are caused due to faulty operation with elevated temperature.