strange problem

Hi,
I got a real strange problem this time. Whenever i use an application that puts heavy load on my NIC and opens a lot of connection (e.g. torrent), my pc shuts down after a few hours (lets say 2 to 8 hours).
It's like unplugging the power cable, no shutdown sequence, no bluescreen, it just turns of. Of course Win 7 is complaining about the improper shutdown when i reboot the system.
Event Viewer isn't helpful: Event ID: 6008:The previous system shutdown at 20:47:07 on ‎11.‎10.‎2009 was unexpected.
and
Event ID: 41: The system has rebooted without cleanly shutting down first. This error could be caused if the system stopped responding, crashed, or lost power unexpectedly.
My first thought was maybe my PSU or some other part of my hardware wants to retire. So i did 4 tests:
1. Running prime 95 (small fft torture test) and boinc (using my gpu for cuda computation) for 24 hours. My system passed this test.
2. Memtest for 24 hours. No errors.
3. Throttling my cpu to 1GHz and 1.1 Volt (instead of 1.4 Volt when overclocked to 2.7GHz) and with only the torrent app running. My system didn't passed this test, it went down after about 3 hours.
4. I ran torrent in a virtual machine (win xp 32bit). My system is stable for more than 24 hours now, so i guess it may be a windows 7 related problem

I really don't know what to do next. Of course i tried different torrent clients. Also note, torrent is just an example. Every application behaving similar leads to the same symptoms. "Normal" down- and uploads are no problem (ftp for example).
Probably important: My NIC is integrated on my mainbord. I didn't mess around with the drivers:
NVIDIA nForce Networking Controller
Driver Provider: Microsoft
Driver Version: 1.0.1.211
Any ideas appreciated,
soulsick

Sounds like the mobo is shutting you down for overheating. That is exactly what it would look like.

soulsick


Utorrent and 64 bit has shown some problems but I have never heard of a drop dead shutdown. Only thing I have ever seen cause that was heat what nic/wifi and router

ken

Keep Core Temp ( Core Temp ) running and watch the temp. If you minimize the Core Temp window, you have the temps in the notification area. You can even adjust the color of the gages, because those dark red colors are not very visible - use yellow.

Also look at event viewer.

Hey guys,
Thanks for responding. I forgot to mention, of course i had an eye on temperatures. I use everest for monitoring.
Highest temps were:
CPU Cores: 59°C, PWM: 60°C, GPU: 52°C (all during my prime+cuda test)
Mainboard 42°C, whenever my NIC is in heavy use (sadly i dont know where the sensor is located)
I dont think 42°C is too much, but on the other hand my mainboard is passively cooled and im not sure about the sensor location.
Hm, i could borrow a nforce4 board from a friend, so u think its worth the effort to completely take apart my system? Any way to make sure the problem is motherboard/NIC related?
Edit: the 2 events i posted above are the only ones related in event viewer, they occur everytime my systems dies.