here's hoping you guys can help me. This is a homebuilt system I put together myself. This is the build (I got a 7870 instead of a 7850, though): http://pcpartpicker.com/p/eZgE
This error has happened to me 5 times now. Once a couple of months ago (Oct 20th), again a month later (Nov 18th), then 2 months later (Jan 28th), last week(Feb 26th), and again now (March 3rd).
What I was doing each time:
First time (Oct 20th): I was browsing Starcraft 2 menus (not actually in a game) while listening to a Youtube video.
Second time (Nov 18th): I was in the middle of a Starcraft 2 game. A 2v2 late-game, so there was a lot of units on the field (maybe more stress put onto computer?)
Third time (Jan 28th): I was in the middle of playing Rocksmith. I was able to reproduce the error a couple of times that night: it all kept happening while playing a song in Rocksmith.
Fourth time (last week): I was in the middle of a Starcraft 1v1 match. It was late-game, so there was a lot of units on the field.
Fifth time (right now): I was playing Starcraft's campaign, and alt+tabbing to browse the web at the same time.
I've googled the issue, and it seems to match Situation #3 here: http://support.microsoft.com/kb/2028504#method3
The error is definitely Kernel-Power Event ID 41. Confirmed via Windows 7's Event Viewer.
The computer never actually restarts, but it seems to hang. Hitting Caps Lock gives no reaction (Caps Lock LED doesnt light up), plugging things into my USB ports (like my iPod) gives no reaction. BUT you can still hear the fans chugging, and the Power LED is still on. So I end up hitting the reset button.
Here's the things I've ruled out:
BIOS update/drivers - I've updated all to the latest BIOS, and updated all the drivers for my computer.
Monitor - not my monitor, since hitting Caps Lock gives no reaction
Overclocking - I'm not overclocking
Memory/RAM - not my RAM. Running Memtest gave 0 errors after 6 passes. I've left it running today, but I doubt I'll see any errors when I get back.
Overheating - Shouldn't be. I've run some Starcraft games and Rocksmith with HWMonitor on in the background. The maximum temps look normal (55C max temps on GPU, normal temps on CPU). Also, IIRC, when my laptop overheated back in the day, everything shut off (which isnt the case here, where i still hear my fans working and Power LED is still on)
Audio drivers - After googling the issue, some people said updating audio drivers/disabling audio drivers solved the issue for them. I've disabled my MOBO's on-board audio completely, so this isn't the issue.
IntelBurnTest - I've run it on Very High stress setting a couple of times before. No errors: the CPU setting/RAM settings should be fine. After narrowing these down, it logically appears to me to be the Power Supply. I have a question about this, though.
Question: If a PSU is faulty, how often does it fail? Does it fail only when under high stress? This seems odd, because it runs perfectly fine in non-gaming situations, so I get the impression I wouldnt even know about this issue if I didn't play games. It's not that my PSU isn't strong enough. The rated 650w should give me around 100w overhead.
Any ideas guys? Any troubleshooting? I obviously want to be more sure that it's actually the Power Supply before I go out and buy a new one.