Sign in with
Sign up | Sign in
Your question

Complete crash without blue screen, fairly new CPU

Last response: in CPUs
Share
December 24, 2011 6:03:50 AM

Hi everyone,

First off, for the tl;dr-ers, a summary:

- Upgraded from duo-core to quad-core 3 weeks ago
- Within last 2-3 days, crashes without blue screen while playing Skyrim, browsing, and safe mode virus scan; temps only in 30s-40s C (heat unlikely to be culprit)
- No major installs, updates, or deletions in last 5 days

I recently upgraded from a Duo-Core AMD Athlon II x2 255 3.1 GHz to a Quad-core AMD Phenom II Black Edition 960T 3.0GHz. Things seemed fine, the stock fan ran a little loud but I don't meddle with overclocking or anything like that so it didn't bother me enough to look into a new fan. Recently, though, my system will completely crash -- no BSOD, and therefore no error codes in Event Viewer to use to find the culprit -- and require a hard manual reboot. There's no issue with POSTing or logging into Windows; the problem has thus far happened when playing Skyrim, but even as simply as playing a low-intensity game on Facebook and even just by downloading a file and running a virus scan at the same time (in safe mode, no less). I'm leery about flashing the BIOS because a) I'm too paranoid about bricking if I don't need to do it, and b) CoreTemp tells me the system will only max out in the high 30's-low 40's C, and I've had FAR worse with a Dell Laptop I used to have. I would be shocked if it were an overheating issue, because 99% of the time I've had a BSOD to signal that sort of problem. I doubt it'd be a hardware issue in general, unless it's stemming from the new CPU -- everything's ran literally without a hitch until now.

My next step is probably going to be a complete reformat, since I've had the computer a little over a year (although the mobo, hdd, and dvd-rom drive are the only original components still in the build) and figure it's probably getting to be about that time anyway, especially since I do a lot of monkeying around with file management and modding games.

Anyone have any advice?

PC Specs:

-ASUS M4A78LT-M LE motherboard
-Quad-core AMD Phenom II Black Edition 960T 3.0Ghz CPU
-XFX ATI Radeon HD 4850 GPU
-XFX 650W PSU
- PNY Optima 8GB (2 x 4GB) DDR3 RAM
-500 GB HDD
-Windows 7 Home Premium 64-bit

Thanks in advance, I'll update if anything else occurs.
a b à CPUs
December 24, 2011 6:44:59 AM

Try to reformat and do a clean instal and see whats hapen after that.
m
0
l
a b à CPUs
December 24, 2011 6:46:11 AM

^^ if all else fails a clean install couldnt hurt...
check device manager under cpu's
m
0
l
Related resources
a b à CPUs
December 24, 2011 7:13:36 AM

it sounds like a black screen reboot which is a power supply problem (psu removes power because it cannot provide a clean enough supply).

try another psu in the system, if you can.
m
0
l
a b à CPUs
December 24, 2011 7:18:29 AM

pauls3743 said:
it sounds like a black screen reboot which is a power supply problem (psu removes power because it cannot provide a clean enough supply).

try another psu in the system, if you can.


that would suprize me bc XfX is a quality company... but it couldnt hurt
m
0
l
a b à CPUs
December 24, 2011 8:58:29 AM

even quality companies have their off days, which is where warranties and RMA comes in.
m
0
l
December 24, 2011 2:51:02 PM

I'd be really surprised; unless the thing is faulty, the new CPU was just a change from 65W to 90W, and I'd upgraded from a 400W PSU to the 650W to accommodate the new graphics card and didn't have any issues. I know that PSUs fail more commonly than most people realize, but would that ~25W change really result in frying the PSU?
m
0
l
a b à CPUs
December 24, 2011 5:19:22 PM

I'd be surpprised if it is a faulty psu as well, but it does happen.

A few months ago I upgraded the cpu and gfx card in my HTPC when it started black screen reboots. The load was still well within the capacity of the psu. After replacing the psu, I haven't had a black screen reboot since. The change was from a 300W FSP to a 450W Silverstone. The system peak load was only ~180W, not a faulty supply as such, just a big enough imbalance in the loads on the psu to cause it distress.

I also blew up a Jeantech 650W (it was an old one) at the start of this year. It powered a Phenom II x4 955 BE (stock speed) with Radeon 4890.
m
0
l
December 25, 2011 8:35:13 PM

Well, couple of updates. Merry Christmas to those that observe; happy Sunday to everyone else!

I ran the computer for about 1.5 hours yesterday before it black screened; this time, however, the thing spontaneously rebooted rather than simply crashing. As it loaded up the second time, it froze shortly after startup with the screen awash in a colorfully-jumbled image. I just fired it up about half an hour ago and ran the Windows Memory Diagnostic, during which the system froze 42% of the way through the first pass and had a comment about a hardware issue being detected.

EDIT: This means I likely won't be reformatting; I'd assume this would limit it to the power supply, CPU, or RAM.

Went in and reseated the RAM in addition to cleaning out the dust and checking all my connections. I've been up and running for about ten minutes at this point and will probably stress test after I finish this post.

I found out from my brother that it had originally blue screened before it began to black screen. I went into Event Viewer and went to look at the times which the computer failed (in fairly rapid succession) and, although I didn't find a BSOD error code, I found dozens of DVD_OV events, labeled simply with "UVD information", immediately before the system crashes. Does that have any significance to anyone?
m
0
l
December 26, 2011 11:51:41 PM

I don't know if anyone will respond to this at this point, but I wanted to update one last time in case someone on Google is looking for the same issue (as that's helped me a few times in the past).

After Windows MemDiag came up with issues, I booted with Memtest86+ from a CD-R. Within a half hour of the first pass I had accumulated over a million errors with both sticks in. I arbitrarily pulled one stick out, ran it again, and came up with another bulk of errors during the first pass. Replaced the first stick with the other, same issue.

I moved the same stick used in the third test over to the DIMM-B slot and ran the test. It went through 2 passes without a hitch. (Yes, I know you're supposed to go through at least 7 to be sure, but I was sensing a significant difference here). Tried another two passes with the other stick, no errors. So changing slots changes the error count from over 1,000,000 in less than half of one pass to absolutely none in two passes. I think we have our culprit; I don't think the mobo is under warranty anymore, so it looks like it's time to shop the after-Christmas Newegg sales for a new one. Thanks for everyone's suggestions.
m
0
l
December 27, 2011 12:43:19 AM

I've been up and running for about ten minutes at this point and will probably stress test after I finish this post.
m
0
l
!