Computer broke...whyyyyyyyy

[DanGWanG]

Warning:  This post is long, so I do appreciate anyone reading the whole thing and giving me some tips on how to resolve this.  It takes a while to explain the problem because I've tried many ways to fix this to no avail.

I built a computer back in October 2010, so it's still fairly new.  Computer worked fine until recently.  It was Dummy OC'd (cuz I'm lazy) for a bit, but I turned it off a while ago, and everything has been at stock speeds since.  I never touched the voltage.

For reference, specs are:

Windows 7 64-bit
Core i5-750
EVGA P55 Motherboard
G.Skill 8GB (2x4GB) RAM 1333
Palit GTX 460 1GB
7200 RPM HDD
Corsair 750w PSU

After installing Rift and Battlefield Bad Company, the computer started glitching like nuts with random artifacts on the screen and computer freezes.  So I tried loading up SC2, and I'd see artifacts and slow/bad performance.  Immediately, I thought it was the graphics card and/or RAM.  Swapped out the graphics card with a good one from another computer, and artifacts are no longer there....awesome.  So I'm playing Rift for a bit, then my display driver crashes within Windows and my computer restarts.  At this point, I'm thinking there is definitely something going on with my graphics card, so I re-tested both cards in another computer to find that its working fine using the same driver.  Both tested graphics cards are GTX 460s.

At this point, I'm thinking it must be the driver or RAM.  I reformat to completely wipe the driver, reinstall Windows and the driver, reinstall Rift and SC2 and start playing.  Rift works perfectly fine for at least 6-8 hours, but I load up SC2 and my system completely starts bogging and the inevitable blue screen appears along with a restart.  This time, and the following few times going forward when this happens, the screen is stuck in the BIOS startup screen and the Windows login sound happens.  So apparently, it went into Windows but left my screen stuck at BIOS splash.  Makes no sense.

Now, I'm thinking it definitely has to be the RAM.  Ran Memtest86 overnight, no problems.  Ran Windows diagnostic test for RAM, no problems.  But I still think it's the RAM, or something to do with the RAM channel.  So, I take 1 stick out, and test both sticks individually while running Rift.  They both work fine on single channel for a couple days of playing Rift.  During this time, I ordered a new set of RAM just in case, and when that arrived I RMA'd the original set.  The new set is in, Rift appears to be working fine for a few hours, then BOOM...computer restarts after a couple hours.  I get back into Windows, and start browsing the internet.  Computer seems stable enough when you're not really doing anything resource intensive, such as browsing internet or typing a paper.  At it's current state, when I load up something like Youtube or any game, the computer display driver crashes and then is typically followed up by a restart.

At this point, there aren't many components left to test with the tools I have available.  There's the motherboard, the PSU, and the CPU.  So, I ordered a new motherboard and that comes in today.  Before I completely disassemble my computer and reassemble it with the new motherboard, are there any suggestions that you fellow GHers could provide?  In my 13 years of computer building as a hobby, I have never run into such a troublesome issue.  I know that the Palit GTX series cards run hot, but I've always kept an eye on it.  I've never seen it go over 70 C, which is an acceptable temperature for the GTX series cards.

Please help me from killing myself.

[.XL]

Please help me from killing myself.

http://www.reddit.com/r/suicidewatch

[DanGWanG]

As I mentioned before I keep a spare power supply around to rule out such mysterious crashes.    

Although in my AssRocks Q6700 build I did last month to replace my burnt out DFI Lanparty (that was easy to debug using my nose) I thought for sure it was a flakey BIOS creating long boots, odd resets, and general wierdness but ended up being a loose IDE cable (old DVD drive I had plugged in).

Damn computers.

Maybe I should check this...

what PSU are you working with?

if you don't see issues show up when doing CPU stress testing, can you isolate it when doing GPU stress testing?  if it occurs on both GPUs, but only in this system, i'd swap out the PSU first.

I have a very similar build using a Gigabyte P55 board and an 5770 GPU and 2 drives.  I'm using a 650watt power supply, and near it's 12v threshold when over clocked under max load.

I'm using a highly rated Corsair 750w PSU.  I haven't done any recent overclocking, and the PSU previously was able to completely handle 2x GTX 460s in SLI with no issues.

[typo]

first check the psu voltages with a dmm. if they are ok, clear cmos and reflash latest bios. this assumes you know how to do this stuff. which you appear to.

the only other explanation is a bad capacitor on the mobo. or bad power regulator. if the machine actually posts it is very doubttful the cpu is dead.

[Lpb45]

I think everyone has hit it right on the head so far it sounds like eithor the psu or the motherboard. I am guessing because you did not mention it that you have not flashed the bios recently for your motherboard?  Have you tried rotating what 6 pins you are plugging into the power connectors on it?  Check the voltage like typo said, check all the little bull**** stuff.  You have been at this a while though Im guessing you have already done all that. They have a list going over at hardforum about 4 series and some 5 series nvidia cards that are having problems with certain motherboards, gonna take a look in a bit.  Will let you know

[Lpb45]

Well there is alot of 460's on this list but I do not know exactly what motherboard you are using.  I would give it a read and see if it matches up with your problems.

http://forums.nvidia.com/index.php?showtopic=187401

[keyb_gr]

first check the psu voltages with a dmm.

A scope would be even better.

[DanGWanG]

Firstly, thank you all for the suggestions.  If I was rich enough to give you all a keyboard as payment for your help, I would, but I'm not rich so..I'm sorry!

first check the psu voltages with a dmm. if they are ok, clear cmos and reflash latest bios. this assumes you know how to do this stuff. which you appear to.

the only other explanation is a bad capacitor on the mobo. or bad power regulator. if the machine actually posts it is very doubttful the cpu is dead.

@typo - Clearing the CMOS and reflashing the BIOS was the first thing I tried, but unfortunately that did not work.  I didn't get a chance to check the PSU voltages, since I don't have a meter at my new place.

I ended up just doing the entire mobo transplant, and now everything seems to be functioning normally.  Will take a few days to thoroughly test, since there could still be a problem (with the PSU).

I think everyone has hit it right on the head so far it sounds like eithor the psu or the motherboard. I am guessing because you did not mention it that you have not flashed the bios recently for your motherboard?  Have you tried rotating what 6 pins you are plugging into the power connectors on it?  Check the voltage like typo said, check all the little bull**** stuff.  You have been at this a while though Im guessing you have already done all that. They have a list going over at hardforum about 4 series and some 5 series nvidia cards that are having problems with certain motherboards, gonna take a look in a bit.  Will let you know

@Lpb45 - Yeah, I've heard of the issues of incompatibility which might have been the case.  After replacing my motherboard, everything appears to be working properly.  The only things I can think of right now without doing additional testing is that either the previous motherboard was bad or there was a graphics card/motherboard incompatibility issue.

[typo]

i bet it was bad capacitors on the mobo or a bad power regulator as i said. those are common failures on mainboards. well, not common but it is what goes wrong if anything does.

a scope? if it is working properly it is full of noise lol. look at the rise time on computer psu's. that is why you can use a $100 ups lol. i have one heck of a scope but i do not need to know what i already know about computer psu's. hey, they work.

[JustCallMeCrash]

Call me a necromancer, cause I'm gonna resurrect this thread.  What puzzles me about this old thread is that no one suggested he try a different hard drive.  As it turned out it was the motherboard, but I'm still surprised this never came up as that's been my problem MANY times.  HDD being one of the few moving parts in a system, they're usually the shorted lived.
Just my pennies.

[flyball]

Call me a necromancer, cause I'm gonna resurrect this thread.  What puzzles me about this old thread is that no one suggested he try a different hard drive.  As it turned out it was the motherboard, but I'm still surprised this never came up as that's been my problem MANY times.  HDD being one of the few moving parts in a system, they're usually the shorted lived.
Just my pennies.

hard drives dont cause any of those problems

[Internetlad]

any swollen or burst caps on old mobo, those can cause all sorts of wacky problems.

EDIT: just realized the OP posted this like 6 months ago, nvm

[JustCallMeCrash]

From the OP: After installing Rift and Battlefield Bad Company, the computer started glitching like nuts with random artifacts on the screen and computer freezes.

I've had hdds do that before.

[Internetlad]

From the OP: After installing Rift and Battlefield Bad Company, the computer started glitching like nuts with random artifacts on the screen and computer freezes.

I've had hdds do that before.

freezes, i'll grant you, artifacts? unlikely. Maybe unloaded textures is what you're thinking of. There's no reason  I can think of that an HDD would cause artifacts.

[Quarzac]

I had a HDD that stopped my motherboard from posting. It was weird.