BSODs and memtest errors after boot (F3-12800CL9D-4GBRL & 870A-UD3)

[OatmealMu]

Motherboard: Gigabyte 870A-UD3
Memory: F3-12800CL9D-4GBRL
CPU: Athlon II X4 635

I'm at a loss...

I just received and installed memory sticks from G.Skill. I go into the bios and change the timings to 9-9-9-24-2T and the CPU NB to +0.100V to get 1600Mhz. 30 minutes after boot and it BSODs. I test with Prime95 and it BSODs after 20-30 seconds. I then try the built-in memory diagnostics tool and the monitor turns black. Finally, I use memtest and get a stream of errors.

Bad motherboard? Bad CPU? The RAM is new, but all these errors...

[MikeH]

May not be related but refer:

http://www.gskill.us/forum/showthread.php?t=7010

or more briefly can you

set the BIOS, "Integrated Peripherals" page, "Onboard USB 3.0 Controller" option to <DISABLED>

and probably rest other timings to a neutral setting while your at it.

You have the AMD version of the Intel P55?

Sorry if it's a waste of time.

[OatmealMu]

I'm using a Gigabyte 870A-UD3, but this been going on for months so I'll try anything. One new hard drive and 2 new memory sticks later and I don't know what to do.

[MikeH]

The GA-870A-UD3 is, to a novice like me, the AMD cpu version of the P55 mobos for Intel cpu's.
I still believe my suggestion has merit especially because your mobo uses the same NEC USB 3.0 feature and that's where my problem lay.
I spent fruitless months looking on the net for any lead possible. and I only wish someone was pushing me to try things left of field.

[OatmealMu]

I might be a bit early in saying this but...

I changed the bios to it's default settings and disabled the Onboard USB 3.0 Controller. Memory Diagnostics passed, memtest passed, and Prime95 has been running for over 30 minutes (it lasted seconds before). You know, I almost hope this doesn't work because the solution was so simple. I would've been saved weeks worth of headaches had I discovered this possible solution earlier. That said, even if this works out, my 1-egg rating on Newegg stays. I'll think twice before buying a Gigabyte product again until they fix their RMA process.

[OatmealMu]

After an hour of Prime95 torture tests, no BSODs and no errors reported. So far so good. I'll spread this possible solution around.

[MikeH]

You are the first to give feedback to my proposal. I have been bent on getting someone else to try the fix and report back, to show it wasn't just me.

I am further blown away, if it really sticks, because you have the AMD implentation of the NEC USB3.0 versus my Intel P55 version.

Please respond after you've done a cold start and whatever you have to throw at it (without bringing on other issues, of course).

The simplest way of showing you have the problem is to re enable the USB3.0 controller and run the Memory daignostics test from Win7 and look for the "Hardware problem detected" message within the first 20 seconds. Then re disable the USB3.0 controller again and see that windows memory diagnostics doesn't say "Hardware problem detected" again. I did it several times to prove the point.

Many thanks for repsonding. Good luck.

[Tradesman]

May want to contact Gigabyte w/ the info and see if they will work on a patch to their USB Controller

[OatmealMu]

Well, that was short lived.

The BSODs are back. My new HDD is apparently bad. Oh, and the new RAM is giving errors in memtest. Seriously, though, what are the odds of two freshly RMAed components being bad?

All that's left now is to replace the motherboard and cpu, I guess.

[Tradesman]

If checking into RMAing the mobo, may want to check for latest BIOS, and if you already have it, go ahead and reflash anyway. Occasionally it will go out of whack when flashed. Seems odd to have DRAM and HDD problems...sorta leaning to mobo problem

[MikeH]

I apologise for any feelings of false hope I may have caused.

I did point out, if this problem existed in your pc it could be simply proven, in 5 minutes, by:
1). Run Win7 Memory Diagnostic look for "Hardware problem exists" within 20 seconds.
2). <disable> USB3.0 controller in bios
3). Run Win7 Memory Diagnostic look for "Hardware problem exists" within 20 seconds. If the message disappeared then the USB3.0 issue was relevant to your situation. If the message "Hardware problem exists" still arises, the pc may have more than one issue and this may still be relevant.

To be fair, I suggested a possible fix for a memory management issue relating to USB3.0. I would never suggest this fix would solve all problems the pc may have and possible also causing a failure on Prime95 and burnin tests. Those tests, by nature, warn of pushing your pc to the point of failure.

I suggested running WINDIAG v0.4 from floppy or CD to test for memory problems. Especially, that I found it showed the same memory slot (A2) having the errors regardless of which memory stick was in the slot. First big clue it wasn't the memory.

I still stand by my suggestion. Anyone, especially with Intel P55 controller mobos, with NEC USB3.0 implementation who is having memory management type issues could take 5 minutes to see if this sorts the problem. But it's not going to sort all problems.

How many persons actually need/use USB3.0 anyway? I'm wondering why it is enabled by default. If I have a problem, the first thing I'll do is disable anything unnecessary and try to isolate the issue.

[mkw87]

I was having problems very similar to yours. I took the RAM out of dual channel slots, put 1 in each of the diff color slots on my mobo, and so far it's stable. Have not found out why. I also disabled usb 3.0, though that alone wouldn't let my computer post when I had the sticks in dual channel.

The weird thing is it worked for over 12 hours in dual channel, then went nuts.

[OatmealMu]

Small update:

I replaced the motherboard with a new version of the same model. It's been around a week and I haven't had any blue screens. Apparently, one of the sticks was defective, but the mobo seemed to be the root of the problem.

[Tradesman]

Thanks for the update, keep us posted