2

A machine that I own does run quite stable in its current configuration. It has been running with a GeForce GTX 295 for a couple of years. That card used quite a lot of power, produced lots of heat and noise. The card got too hot easily and I had to start cleaning it quite often.

I switched to an older PCI-Express card (GeForce 9500 GT). This ran stable.

I decided to upgrade to a GeForce GTX 650 Ti. The system starts up fine (Windows 7 x64 or Debian GNU/Linux with Xorg+Nvidia's drivers). Quite soon (<5 minutes) after the boot has completed the machine turns off now, though: The machine turns off (including fans) and 1-2 seconds later it boots again (different from a reset, where the fans aren't noticeably turning off in between). This happens faster when I try to enter some game: it's mostly instantaneous in that case.

This card runs an ran stable in another system for months with lots of games that were played. It's unlikely to be broken. Now I'm wondering what components are the most likely culprits. I suspect the following parts in that order:

  1. PSU. It's a Coolermaster M1000 (1000W) which should have no problem if it worked well. I tried the different dedicated PCI-Express connectors and an adapter which used 2 of the 12V connectors to power the graphics card. Even connecting to the ports at which the HDDs are working fine did not help. Still, the PC turning off very much looks like the graphics card's power supply might be unstable to me.

  2. Mainboard. I have tried different PCI-Express slots, the capacitors look fine. But still...

  3. The card or other components: Unlikely. The machine and the graphics card both run fine, even under high loads, just not together.

What do you think, who is the most likely culprit? Could it be the BIOS settings, too? Anything I should look for? I will most likely replace the PSU (cheapest and most likely option), but still wonder whether my guesses are good.

Some specs: 2 SATA-HDDs, 1 DVD-RW, ASUS P6T Deluxe mainboard, i7-920 CPU (2,6 GHz @ 3,6 GHz - but same with default BIOS settings), 24 GB DDR3, two screens at 1920x1200 each attached, current drivers, no crashes even with Prime95 or similar running.

1 Answer 1

2

My best guess would be the power supply as well. Just because it has enough watts doesn't mean there isn't some other problem. Can you get ahold of another PSU just for testing? Another thing to consider is how good of a heat sink do you have on your CPU? If the video card is throwing off a lot of heat, it can be adding to the heat of the CPU. Modern motherboards have sensors that will automatically shutdown the computer if it gets too hot.

1
  • The PSU has been the culprit indeed; of course it may be just about anything else in the general case, but I am happy if my reasoning seems to be fine. May 6, 2016 at 23:39

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .