2

I ordered parts for my computer a few weeks ago.

  • PSU: Corsair CX 430
  • CPU: AMD FX 6300
  • GPU: GeForce GTX 750 Ti
  • RAM: 4GB 1600MHz x2
  • Motherboard: Asus M5A78L-M

When I assembled them, I realized there was a problem when I tried running graphically intense programs - I first ran a game which immediately crashed my computer - the screen went black, as if I had disconnected it but the fans kept spinning until I restarted the PC. I tried benchmarking with OCCT and Kombustor which gave me the same result. I also made sure nothing is overheating, I replaced the RAM and I re-sealed everything.

My initial assumption was, that the GPU is faults, so I emailed the seller, who advised me I try a different (stronger) PSU. I borrowed a cheap 550W PSU (LC Power LC6550) from a friend, and replaced it.

When I tried running games with the replaced PSU, the GPU worked fine. I was also able to benchmark it in Kombustor. It was all great, however, now the same problem occurs when the GPU heats to about 65°C. Screen goes black, fans spinning, etc..

I am unsure what is the problem now - at first, I though it was graphics, but after replacing the PSU I think it could be that.

  • I have never seen the behavior you described by anything more then a system not getting enough power. – Ramhound Aug 28 '14 at 10:57
  • @Ramhound so you suggest is's a PSU problem yet again? – Gregor Menih Aug 29 '14 at 10:12
  • LC Power is crap. Time to get a decent PSU. With your system, no more than 450 W should be necessary. – Daniel B Aug 31 '14 at 9:58
  • 3
    Your system probably never draws more than 175W from its power supply. – David Schwartz Aug 31 '14 at 11:05
  • I agree with most of the answers/comments: first try to get a decent PSU with a stable voltage and enough amp's. Second, you could try to test your GPU's memory (instead of the CPU) using for example this tool. – agtoever Sep 1 '14 at 21:16
4

Are you sure that your system goes haywire at GPU's temperature going over 65°C? Because in my opinion, 65°C is not hot. Hot is when GPU reaches 80-90°C. Even then it works. I believe the problem is somewhere else.

Just to be sure, blow any dirt off of the heat sink. If you feel comfortable you could remove the heat sink and replace the thermal paste. Then check if the problem persists.

  • Yes, I am sure. At first I thought it would shut down randomly, because the GPU didn't heat up to 65°C. After trying to play graphically intense games, it would shut down after a few minutes, so I tried OCCT and Kombuster, and when the temp got to 65°C it shut down. There is nothing in the heatsink, the GPU is 2 weeks old. I am hoping the problem is elsewhere. – Gregor Menih Aug 28 '14 at 23:06
  • 3
    @ItsGreg He’s right though: 65°C don’t qualify as “running hot”, not even remotely. Maximum temperature as per specifications is 95°C. – Daniel B Aug 31 '14 at 9:59
  • @ItsGreg - Have you checked the CPU temperature at the same time? – Niks Aug 31 '14 at 18:40
  • @Niks, yes of course. It's usually below 50°C. Also, I ran GPU Stress test, so the CPU was barely in use. – Gregor Menih Aug 31 '14 at 20:52
0

This problem can be caused by different things. Please make sure the following things are in order

  • Install the latest drivers for your graphics card and motherboard. Download the latest versions from the manufacturers website.
  • Make sure all the power cables are connected well enough (24 pin motherboard, 4/8 pin CPU, PCI-E power connectors for you graphics card etc)
  • Make sure your case has good airflow. If it doesn't, try running tests with the side panel off.
  • The problem may also be caused by the power delivery components on your motherboard. When your graphics card is doing intensive operations, it needs more power. Maybe your motherboard isn't able to deliver enough power to your graphics card (or too much) which will cause the system to crash. This would also explain why replacing the RAM or even the power supply doesn't help (the other power supply might deliver power in a slightly different way, that's why the results are not exactly the same as with your Corsair PSU).

65°C is not that hot for a graphics card and shouldn't cause any problems. I think the problem is with power delivery so check with Asus if you can get a replacement or borrow someone's motherboard to test it out.

Good luck!

0

The PSUs power is enough and it also is a decent one. I compared benchmarks results of your CPU and mine (i5 3570K) and your CPU is consuming ~50 Watt (full load) more than mine. Seen here. My GPU on the other hand is more power hungry than yours and I run my CPU overclocked with a 450 W PSU.

For me there are two possibilities what might cause this.

  1. Your PSU is broken. If your CPU or RAM would be faulty it would crash, freeze or cause a Bluescreen but not shut down completely. If the GPU would overheat you would see artifacts and weird colors before your game crashes -- bun again in this case it wouldn't shutdown.
  2. Your CPU is overheating and shutting down to prevent further damage. In that case you can run Prime95 to stress your CPU without any load on the GPU. If this happens it might be possible that you did not place the CPU cooler correctly.

Not the answer you're looking for? Browse other questions tagged or ask your own question.