computer crashing, no BSOD

[Fayd]

This happened yesterday while playing Mass Effect 2. The computer experienced a crash, I could not recover from it. The mouse froze, and the computer would not respond to CTRL-ALT-DEL.

I rebooted (it responded to a soft-reboot) and loaded up firefox (nightly). Crashed 2 more times last night, same symptoms, while just using firefox. (mouse frozen, won't respond to ctrl-alt-del.)

I tried this morning, used sfc /scannow to see if there were any windows issues. Didn't spot any. I tried looking for smart info to see if there were any unrecoverable errors on my SSD, but can't get smart info through AMD SB750 while raid enabled. (use raid1 for storage)

downloaded and ran memtest86+ to look for memory errors. 1 full pass, didn't spot any.

downloaded hwmonitor and prime95 to look for changes in voltage. loaded hwmonitor, voltages look stable. load prime95, 8 threads (it's an fx8320, stock clocks). crash as soon as i click go.

reboot, load hwmonitor, load prime95, 4 threads.. crash 1 sec after i said go. not enough time to observe voltages.

what the heck is going on here? this computer was solid as a rock 2 days ago, now it crashes like a drunk 2 year old on the freeway.

specs:

Gigabyte GA-970A-UD3
4x4gb Kingston 1600 DDR3 (running at 1333, too lazy to correct it)
AMD FX 8320
MSI R270 twin FROZR (actually not sure it's MSI.. it's an R270 with 2 fans)
Corsair HX520
Corsair Force 120gb ssd
WD black 1tb
Hitachi 2x2tb raid1

Ideas on what's going wrong?

 

[Drummerdude]

typically a freeze is caused by several things. A failing hard drive, a failing/failed cpu, or I've even seen bad ram cause a bad lockup. If yours is freezing hard the second you hit go on prime95, it's very possible your processor went bad. ( overvoltaged, power hit, etc ). Amd's are cheap. Maybe if you have a spare processor to test with, that'd help. Also, if you have a spare hdd, disconnect your main hdd, install your os to spare, try prime95. If it doesn't crash once over several tests, you have a couple possibilities. Could be a failing hard drive, could be severe file corruption ( maybe? ), could be controller driver bad.

 

[Fayd]

typically a freeze is caused by several things. A failing hard drive, a failing/failed cpu, or I've even seen bad ram cause a bad lockup. If yours is freezing hard the second you hit go on prime95, it's very possible your processor went bad. ( overvoltaged, power hit, etc ). Amd's are cheap. Maybe if you have a spare processor to test with, that'd help. Also, if you have a spare hdd, disconnect your main hdd, install your os to spare, try prime95. If it doesn't crash once over several tests, you have a couple possibilities. Could be a failing hard drive, could be severe file corruption ( maybe? ), could be controller driver bad.

I don't overclock nor overvolt. As stated, my memory is even running downclocked. I just don't bother with it enough to justify overclocking.

the computer has been behind an UPS its whole life. the UPS is fine. processors don't typically go bad. it's solid state, nothing changes. so if it went bad (unlikely imo), something else caused it to go bad.

I checked in AMD's raidxpert utility, and it said the smart status of all the HDD's is good. I know that doesn't mean much without being able to check the actual readouts, and i know smart status doesn't mean a whole lot. that's why i ran sfc, to see if there had been a corruption error that took out the windows install.

 

[stinger608]

Hmm, I am actually thinking along the lines of something going amiss with your power supply.
I know Corsair is awesome hardware, but having a 520 watt is a little less than what you should be running and it may have put undue stress on the power supply circuits.

If you have the means, pull all the plugs on the power supply and fire it up to test voltages coming out of the power supply to the main board, cpu plug, and PCIe plug.

If you don't have a way of actually testing the supply, see if you can beg, borrow, or steal a different power supply to test the system again.

 

[Ketchup]

If you are not overclocking, sounds like power supply or CPU to me. Do all of your fans seem to being running normally?

 

[Fayd]

If you are not overclocking, sounds like power supply or CPU to me. Do all of your fans seem to being running normally?

the temps were normal. even still, high temps would cause it to fail after some time, not immediately.

i'm inclined to believe that it's probably the power supply. the existing PS is around 7 years old. I already ordered a new one from amazon. (a seasonic 550w 80+ gold rated)

 

[John Connor]

I had (may still have) freezing problems. At first I thought it was the new BIOS I installed, so I downgraded and still had intermittent freezes. So I updated the graphics card driver and it hasn't froze in a while, although I think it did even after the new driver. Although it hasn't froze in a long time now. I also use alternative drivers for my sound card and maybe that could be it too. Ever since I installed Win 7 I was getting the freezes.

 

[Automaticman]

Are you getting actual BSOD's though? I'd say check Windows Event Monitor for critical faults, but I wonder if they would be logged if the PC didn't actually get to the BSOD screen. Worth looking anyway.

You try unplugging any extra peripherals?

 

[Fayd]

Are you getting actual BSOD's though? I'd say check Windows Event Monitor for critical faults, but I wonder if they would be logged if the PC didn't actually get to the BSOD screen. Worth looking anyway.

You try unplugging any extra peripherals?

they didn't get logged. I Already checked that.

 

[phis6]

This happened yesterday while playing Mass Effect 2. The computer experienced a crash, I could not recover from it. The mouse froze, and the computer would not respond to CTRL-ALT-DEL.

what the heck is going on here? this computer was solid as a rock 2 days ago, now it crashes like a drunk 2 year old on the freeway.

specs:

Gigabyte GA-970A-UD3
4x4gb Kingston 1600 DDR3 (running at 1333, too lazy to correct it)
AMD FX 8320
MSI R270 twin FROZR (actually not sure it's MSI.. it's an R270 with 2 fans)
Corsair HX520
Corsair Force 120gb ssd
WD black 1tb
Hitachi 2x2tb raid1

Ideas on what's going wrong?

Try doing a system restore to a restore point way back 2 days ago before having that problem. You might have installed updates that caused conflict in your system.

 

[John Connor]

Whocrashed can show you what crashed.

 

[Fayd]

Whocrashed can show you what crashed.

only if there's a log and a dump file.

without those, there's no way to know. it's crashing in such a way as to not generate a log or dump file.

 

[John Connor]

only if there's a log and a dump file.

without those, there's no way to know. it's crashing in such a way as to not generate a log or dump file.

Do you have updated graphics drivers?

 

[Fayd]

Do you have updated graphics drivers?

do you have a checklist and are just moving down it?

i have the most recently released non-beta drivers for the GPU. but i HIGHLY doubt the GPU is relevant to the problem. the problem occurs when I place heavy strain on the CPU. I actually loaded back into mass effect after a time and played the game for another 2 mins before making some choice that would require more loading, and it crashed again.

as I said, i'm thinking it's a PSU load problem. the corsair unit in it now is 7 or 8 years old. I ordered a new PSU and it'll be here wednesday. If that doesn't solve the problem, back it goes, and i'll take a heavy look at the capacitors around the CPU. (though i think this board is solid state caps? hmm...) at that point, i'll pull a video card from another system and try it.

 

[stinger608]

I think, like I said earlier and you tend to agree, it is the power supply more than anything else.

After getting your new power supply and installing it I think your problems are going to be solved. I have to think that it is either the 4/8 pin or 24 pin not delivering proper voltage and when it is put under a heavy load it shuts down.

Had this very same issue with a customers system about a month ago. Would run just fine doing most anything until the system was put under a heavy load........Then boom, shut down and restart. Some times it would just shut completely down.

Installed a new power supply and the system was just fine.

None the less Fayd, keep us informed on what happens Wednesday once you get the new power supply installed.

 

[John Connor]

do you have a checklist and are just moving down it?

That was only my second post and in my fisrt I mention how I had freezing problems and I updated the GPU driver and it seemed to fix it.

What freaking list?

 

[Fayd]

replaced PSU, same problem.

tried running prime95 by lowering the thread count to 1,2. in both cases, both threads exit almost immediately with hardware failure.

furmark though runs without issue.

 

[Smoove910]

just going on a whim... but is your ram this?
http://www.newegg.com/Product/Produc...82E16820104169

If so, have you checked the voltage of your ram in the BIOS? I've seen numerous times where someone will buy 1600mhz ram, but the BIOS defaults to 1.5v when it requires 1.65v. When this occurs, the results are like the issues you are describing (crashing without warning, crashing when load is applied, etc).

Might be worth a shot, although if your ram is still running at 1333mhz, it may be a non-issue.

Good Luck!

 

[Fayd]

it is 1600mhz, 1.65v ram that i've set default to 1333 mhz, 1.5v.

I tried changing ram to rated spec, rated voltage, and it failed to boot. maybe I did it wrong. came up with some message about how clock settings in bios caused failure to boot.

i left the clockspeed at default, bumped the voltage to 1.6v, but no change. besides, i've run memtest86+ for a full pass on the ram without issue.

 

[Ketchup]

Have you tried the 'load optimized defaults' option in the BIOS?

 

[Fayd]

Have you tried the 'load optimized defaults' option in the BIOS?

yes, that was among the first things i tried. but since i already run optimized defaults (only thing changed is using raid instead of IDE), as expected, no impact.

 

[Smoove910]

what is your CPU voltage set to? Do you have it set to 'auto', or is it set to the default of 1.35v?

Was just re-reading through your first posting and the part where you said you loaded 8 threads of Prime 95 and it crashes as soon as you hit go interests me. To me, this has always happened when my CPU didn't have quite enough juice. Granted, I'm an overclocker, but the symptoms are similar.

Lastly, what BIOS revision are you running? I know your issues just recently started, but at this point I would weigh the option of updating to the latest if possible.

 

[Fayd]

what is your CPU voltage set to? Do you have it set to 'auto', or is it set to the default of 1.35v?

Was just re-reading through your first posting and the part where you said you loaded 8 threads of Prime 95 and it crashes as soon as you hit go interests me. To me, this has always happened when my CPU didn't have quite enough juice. Granted, I'm an overclocker, but the symptoms are similar.

Lastly, what BIOS revision are you running? I know your issues just recently started, but at this point I would weigh the option of updating to the latest if possible.

it is set to auto, but the default voltage at auto is 1.45v. i imagine under coolnquiet, it's probably down to 1.1. But i'm not having problems with low voltage computing, only when i actually demand something from it.

f7, the latest non-beta bios. (f8 doesn't update anything i need..)

At this point, I'm starting to think the CPU has gone bad. Next weekend (i don't have time during the week), i'm going to pull/swap ram sticks to see if that will fix it. if not, I'll RMA the cpu to amd. if that doesn't fix, it's gotta be the motherboard.

actually one thing I haven't done is try stressing it in absence of windows. I don't think this is a software problem, but it would make sense to test to eliminate the possibility. i downloaded a copy of stresslinux, but haven't burned and loaded it yet. i could do that next.

 

[Fayd]

Finally got around to trying to stress it outside of windows. Downloaded a linux livecd, and ran mprime on it. crashed the same. exact same symptoms. really weird. At least it's consistent, though.

thus far, I've ruled out power supply (though the motherboard may be borking the power conversion), and anything relating to hard drives. (software...) so it must be hardware, central to the computer. I haven't ruled out graphics card (but given it's only when the CPU is stressed, not the GPU, i don't think it's the GPU. I've ran furmark no problem... but the instant prime95 starts up, crash.) So at this point, it's processor, memory, or motherboard. I've already ran a single pass of memtest86+, though. no failures.

 

[Fayd]

For anyone that encounters the same problem and symptoms,

The problem was the CPU. I swapped the old CPU out with another system's CPU, no problems. RMAed the 8320 to AMD and explained what happened. They sent me a new 8350... plugged it in, started up just fine. working right now with 8 threads prime95.