PC shuts off after exiting games

[17seconds]

I have been trying to figure this out for weeks now. Randomly in October I started to experience a simple problem: My PC shuts off about 2-5 seconds after exiting a game. The PC attempts to restart, but usually suffers through several attempts where it shuts off just as the monitor is supposed to come on. I can then always get the PC to turn on by switching off the power supply using the master switch on back.

Temperatures seem normal for all components, with the CPU usually at 34c, and in the mid-40c range after gaming. Strangely, I can run MSI Kombustor (Furmark) with no problems without the PC shutting off. This happens with all my games, though: Unreal Tournament 3, Dragon Age: Origins, Half-Life 2, Metro 2033, Mafia 2, Fallout 3...

The only thing I have noticed, it may be normal, is on CPU-Z, at idle my CPU has a multiplier of 9 (x133 mhz) = a speed of ~1200 mhz. I am able to see the multiplier go up to 22, 24, 26, etc. as the Turbo mode kicks in.

Some remedies I have tried:
-Exchanged (RMA'ed) my PSU - Corsair 850HX
-Exchanged (RMA'ed) my video card - GTX480 for a GTX580
-Changed the thermal paste, reseated the heatsink - Corsair A70
-Updated my motherboard BIOS to the latest
-Updated and rolled back video card drivers
-Turned off all overclocking

I'm running out of components to rule out, and I think I'm left with the CPU or Motherboard (see my setup below). Are there any tests I can run? Again, it's just when exiting games.

Thanks for your help.

 

[zergesys]

The memory may be another thing to check. Just use memtest for that, and try swapping out sticks.

Also consider running 3D mark. The way it gauges graphics and CPU performance may reproduce the problem for you to help narrow things down.

As for the clock of the CPU, that's indeed perfectly normal. It's just "stepping". A system that dynamically changes the multiplier and voltage according to loads to help with things like heat and power use.

 

[17seconds]

The memory may be another thing to check. Just use memtest for that, and try swapping out sticks.

Also consider running 3D mark. The way it gauges graphics and CPU performance may reproduce the problem for you to help narrow things down.

As for the clock of the CPU, that's indeed perfectly normal. It's just "stepping". A system that dynamically changes the multiplier and voltage according to loads to help with things like heat and power use.

I forgot to mention that I can also run 3D Mark Vantage with no problems. I will try Memtest. At this point I am looking for good diagnostic and hardware monitoring software to use.

 

[zergesys]

SiSandra may be able to help you out. It's a very well fleshed out suite of diagnostic utilities.

 

[17seconds]

Okay, I ran AIDA64 to stress test the CPU. The PC shut down just like it did when I was playing games. The temps got into the mid-50c range. It also took a few tries before it was able to complete a boot up.

So, I'm thinking it's a CPU problem, but I'm not sure what the solution is. There must be something in the BIOS I can configure to change the way the way the CPU responds when stressed.

 

[17seconds]

From when you are facing this problem?I mean did you install any new hardware or software.Can you tell me how your pc restarts?I mean does it completely shut down like someone unplugged the power cord.Did you tweak any ram voltage?Run pc with one stick of ram.Try re-sitting your cpu.

And you must be tired after fighting that long flamewar with physco eyefinity.lol

i'm pretty sure that it started happening for no reason. I can't think of anything that I added or changed right before it started happening. It's just as if the power plug was pulled out, the screen goes blank, fans shut off, and the PC starts, or tries, to reboot after a few seconds.

I have definitely found that my system doesn't like running Prime95 or anything that stresses the CPU. I tried the test in Prime95 that put minimal stress on the RAM, and the shutdown problem occured. Temps are not getting above the mid-50c's.

I am ready to re-seat the CPU as I have changed thermal paste a couple times over the past couple months. I also think there might be a BIOS setting involving power management that may help, but I don't know which one.

Is Eyefinity's thread still going? I added some quotes and article links, just to "balance" out his rage, but then unflagged the conversation to stop following it. That stuff is just ridiculous how some seem to be so emotionally invested in computer hardware.

 

[17seconds]

Okay, I finally got it figured out, and the winner is....
a faulty motherboard.

I was able to replace the motherboard and all the problems went away. Someone suggested that the problem had to do with the system power. Since it wasn't the PSU, the next logical culprit was the motherboard. Hopefully, the RMA process with MSI takes less than two weeks, before I have to return the temporary motherboard.

(On a side note, I have gotten so good with taking components out and replacing them that I can now take out and replace a motherboard and have the machine running in about an hour.)

 

[17seconds]

Best answer selected by matto17secs.