Sign in with
Sign up | Sign in
Your question

Periodic crashes on new system with GA-X58A-UD3R motherboard

Last response: in Motherboards
Share
September 8, 2010 5:09:44 AM

Hello,

Recently I got a system built with the following specs:

Gigabyte GA-X58A-UD3R motherboard
DDR3 6GB 1600 MHz Patriot memory
ATI Radeon HD 5570 graphics card
Intel Core i7 930, 2.80GHz
Windows 7 (64 bit version)
650W Antec Earthwatts power supply

My issue: the system has been crashing periodically, with the monitor (Acer, 23") displaying an odd checkerboard pattern before everything freezes and Windows restarts. The graphics card drivers appear to be up to date. One thing I noticed is that the BIOS shows the memory being recognized as 1066Mhz, lower than the memory's rated frequency, but I didn't see a reason why more conservative memory timings would cause this issue. I did notice that Gigabyte's website did not mention that this particular memory frequency (1600Mhz) is supported, and was wondering whether I should try DDR3 1333MHz modules. Would appreciate any thoughts or suggestions on diagnostics I can do. (The BIOS version is FA, but I wanted to exclude other possibilities before upgrading the BIOS).

Thanks!
a b V Motherboard
September 8, 2010 5:48:23 AM

It sounds more like a graphics card or psu issue, but to eliminate memory:

Run Prime95 stress test, and make sure "check for rounding errors" is on. If you can run that for 30-60 min without a crash and without core temps going too high (use CPUID's Hardware Monitor to watch temps), memory and cpu are cleared.

To stress the gpu, run Furmark while watching gpu temps.

Let us know what happened.
m
0
l
a c 717 V Motherboard
September 8, 2010 2:21:21 PM

Q - How many HDDs & SSDs are running? The reason is that MOBO has a problem with SSD + HDD running on the Intel SATA2. As I recall the FA is the current non-beta BIOS so don't flash.

Clearly, unless your DDR3 has a bad stick your DDR3 is not the cause being under-clocked. If you post the exact P/N of the DDR3 I can suggest settings to properly OC the Patriot memory.

Q - Is there any "particular" time when the rig crashes (e.g. 1 game, data processing, etc)?

To R/O DDR3, create a Memtest86+ bootable CD/DVD using the ISO. Yes, this could be the GPU or simply the GPU displaying garbage prior to a shutdown. However, my money is on the MOBO itself and/or some other configuration(s).

Next run MSCONFIG -> select Diagnostic Startup + Restart; see if the problem is indeed repaired. Refer to this link - http://www.sysinfo.org/startuplist.php

Also, in some instances Adobe Flash has been know to have similar odd behavior so for now uninstall + Test to R/O.

Testing & Benchmark Tools: Prime95, CPU-Z, CoreTemp 0.99.6, Memtest86+, & 3DMarkVantage

1. Prime 95 as a "Stress Test" [overnight.]
link - http://www.mersenne.org/freesoft/

2. CPU-Z for Settings Summary
link - http://www.cpuid.com/softwares/cpu-z.html

3. CoreTemp 0.99.6
http://www.alcpu.com/CoreTemp/

4. Memtest86+ to check memory corruption
link - http://www.memtest.org/

5. 3DMark
link - http://www.futuremark.com/benchmarks/3dmarkvantage/intr...

Good Luck - post back!

After the above is R/O it gets more complex to determine the cause. {PSU,GPU,etc}
m
0
l
Related resources
September 9, 2010 3:03:38 AM

Thanks for the suggestions!

To first question- no SSD, just one HDD (Western Digital Caviar Black, 1.5TB).

P/N of DDR memory is PSD36G1600KH

To second question- crashes have occurred at several times, when running Windows Easy Transfer, when running Windows Backup to back up to external drive, during Windows update, and one additional time when I think only application open was a web browser.

Also, for completeness, I was not able to get Windows Backup to complete a backup of the HDD to an external drive- it suggested running CHKDSK. I did run CHKDSK on the C: drive and the external drive and it came up clean. Not sure if the issue is even related but thought I would mention it.

I'll try the things you suggest- thanks again!

m
0
l
a c 717 V Motherboard
September 9, 2010 3:27:05 AM

Boy that was fun to find the CAS settings, they're 9-9-9-24 and the Memory Multiplier will need to be set to 12 {BCLK 133.33 X 12 = 1600 MHz} - but that can wait.

Not to add to the to-do list, but you probably should run both options on Check Disk {scan + fix}. If you mounted/dismounted an attached HDD {back-up} that wasn't in AHCI you can corrupt all sort of things. If you have any password Windows Login temporarily disable while trouble shooting, and updating Windows. In addition, I would try Safe Mode + Networking for updating.

However, my money is on -> Next run MSCONFIG -> select Diagnostic Startup + Restart; see if the problem is indeed repaired.

Good Luck - Let us know!
m
0
l
September 11, 2010 2:45:34 PM

So, I think I made some headway- here's an update.

- ran memtest86+ (a nice program, btw)- no fails after 2 cycles.
- ran Prime95 overnight as CPU test (>6 hrs). Monitored temp and no core got higher than 66 degrees during test.
- downloaded 3DMark program and ran- everything seemed to run OK. 3DMark score looked roughly in line with the few other scores I saw with this graphics card.

Restarted with msconfig and ran Diagnostic Startup - used machine for several hours with no evidence of a crash. Re-enabled startup items and a crash occurred within about 10 minutes, so now suspecting one of the startup items is at fault. Most of them look to be fairly innocuous (referring to the sysinfo.org link) - I imagine process of elimination will the only way to identify the one at fault but would appreciate any other suggestions.

Windows backup still having issues with external drive; ran chkdsk on backup drive- right now running chkdsk on the C: drive to see if it finds anything.

Thanks again for the suggestions!

m
0
l
September 11, 2010 2:46:53 PM

Twoboxer said:
It sounds more like a graphics card or psu issue, but to eliminate memory:

Run Prime95 stress test, and make sure "check for rounding errors" is on. If you can run that for 30-60 min without a crash and without core temps going too high (use CPUID's Hardware Monitor to watch temps), memory and cpu are cleared.

To stress the gpu, run Furmark while watching gpu temps.

Let us know what happened.



Thanks for the suggestion on Furmark- a good program- if I ever add a fancier graphics card it would be interesting to try again!
m
0
l
a c 717 V Motherboard
September 11, 2010 5:50:53 PM

So how's the Back-up coming? Is the problem fixed yet?
m
0
l
September 12, 2010 6:57:55 AM

not yet. I ran chkdsk with both options {scan + fix} checked on both the C: drive and the external drive, and both came up without any errors. Ran Windows Backup again and got the same error code (0x8078012D) as previously. I am able to write individual files to the drive without issues. One observation is that the drive light now flashes continuously at ~1Hz- which is supposed to indicate the drive is reading/writing- except this now is occurring even when nothing is being actively written to the drive.
m
0
l
a c 717 V Motherboard
September 12, 2010 3:07:41 PM

Okay, this is outside my expertise, I found what appears to be a valid and related "possible" fix. It is similar in that it is a Diagnostic approach as I suggested earlier and has additional MS links relating to the issue {all in one}.

If this corrects the problem I'd really like to know - Thanks!

See Gaurav Prakash / Microsoft Support - http://social.answers.microsoft.com/Forums/en-US/w7repa...

Good Luck!
m
0
l
September 22, 2010 4:05:24 AM

FYI- both issues are now solved! The crash issue was the graphics card- got it swapped out and new card no longer has issue.

The Windows Backup isssue was a conflict with a Gigabyte application (SmartRecovery) that apparently deletes some files on the hour. Windows Backup failed every hour, on the hour, which I hadn't noticed before. Some guy on one of Microsoft's support sites had encountered the exact same issue.

Thanks again for the pointers and helpful advice- learned a great deal in tracking this down!
m
0
l
a b V Motherboard
September 22, 2010 4:35:10 AM

Great!
m
0
l
a c 717 V Motherboard
September 22, 2010 5:01:11 AM

glcaan99 said:
FYI- both issues are now solved! The crash issue was the graphics card- got it swapped out and new card no longer has issue.

The Windows Backup isssue was a conflict with a Gigabyte application (SmartRecovery) that apparently deletes some files on the hour.


Yeah, NEVER use ANY of Gigabytes Utilities or Apps. I am surprised that the MSCONFIG recommendation I had didn't stop the Gigabyte Utilities/Apps with that problem not withstanding the GPU issue.

Glad to hear it's fixed! :) 
m
0
l
a b V Motherboard
September 22, 2010 5:24:48 AM

Never use two apps that use "unused time in the background" together is really the issue.

You're very lilkely to get the same failures using, eg, Norton (background scan) and Memeo Backup software. And probably either of those would crash windows backup, too.
m
0
l
!