Solved BSOD very random - All tied to ntoskrnl.exe as crash address

My custom made PC has been plagued with very random BSODs for over a week now but all crashes logged show the crash address as ntoskrnl.exe. For nearly all the crashes (except this very latest one while I was setting up my account here and files to share with you) happened while running a game, 1 Firefox window, and 1 Chrome window with a streaming TV channel.

I recently removed my NVIDIA GTX 570 to upgrade to the GTX 680 (yes, full driver flush and reinstall) and since then and possibly tied to the latest drivers that were updated (301.42) I've been having issues. I suspected it was just a bad card (which I got directly from NVIDIA after working with them at PAX East).

The card was removed and I purchased a GTX 560ti to test things but I've crashed twice today already.

When I ran memtest86 after the first few crashes, my 16 gigs of G-Skill ram passed all tests but the block move (test #7). I've read that that may be due to the latest build of memtest86 so I have not suspected my RAM to be the issue. However, now that I've pretty much determined it isn't the video card, I am completely stumped.

All drivers have been properly updated (bios, chipset, SATA, etc.) and I was hoping someone could spot what is going on with the files I've uploaded and according to the instructions in your sticky. Please advise what other info I can provide, I'm new here.

As an aside, when I had my GTX 570, I never had issues and could run the same programs as I have been during this phase of trouble. Also, I have been informed by some co-workers that I have some conflicting issues with the AMD Catalyst Control Center and the fuel service. When I was updating my SATA and chipset drivers, I was led to believe I had to have the CCC to update them but now that's causing rampant errors which are logged in my Event Viewer.

If anyone sees anything beyond that, please do share, I'm still troubleshooting these issues as I type this.


EDIT: AMD CCC, USB filter, and APP SDK have been uninstalled.

UPDATE and new files attached.

The GTX680 is back installed since I BSOD'd with the 560ti test card. NVIDIA drivers fully removed and reinstalled clean.
RAM tests run with memtest86 4a and memtest86+. Both showed errors on the block move test (both versions) and only with that test.
RAM CAS settings in BIOS were set to auto and had things set to 11-11-11-28, G-Skill for my model recommends 9-9-9-24 at 1.500V. I made sure these were corrected in BIOS.

Once all retweaked and set, I loaded back up and ran Diablo III with no other windows open other than a Google Talk window to my friend who has a degree in computer tech. I BSOD'd again after about 20 minutes of game play with the blue screen suggesting I locked up on the dxgmms1.sys file but in Blue Screen Viewer, it states I failed with ntoskrnl.exe.

Any help--please--is deeply appreciated.


EDIT:

Steps I've taken - sfc /scannow (100% ok), sfc /verifyonly (100% ok), Device Manager (no errors), hard disk full scan (no errors), all drivers checked and updated as needed including MB and CPU.

Suspicions - Latest NVIDIA drivers are broken since this began to happen around the time I updated to them. Or, RAM may be having issues (checking 4 sticks is going to suck) as G-Skill's site does not list my MB as being approved.

Anyone, please?

Hello and welcome to Seven Forums

Now I'm not a BSOD expert but I would like to say, your assistance will come. Our BSOD specialists are being bogged down with requests for help at the moment and as a result are struggling to keep up with demand. They will get to you but please be patient.

Stephen.

Recommend removing your Gigabyte utilities. Your power utility software is out of date, and when it is out of date, it is known to cause problems on some systems. EasyTune6 is also an issue when it is out of date.


If that above do not resolve the crashes: Your HTO CLARO Audio driver is out of date and probably causing conflicts with both your graphics card and your network drivers. Recommend updating it from HT Omega - Driver Downloads

Thanks, writhziden.

Okay, so in all the muck (especially the update zip with the latest BSOD from yesterday and my "reset" machine), the errors you saw were for the added programs that came with my MB? They shouldn't be running on start-up, is why I ask, and I normally don't use them at any time. The only programs I have pulled up were Gigabyte's @BIOS to check for an update and the ET6 to monitor voltage settings and what-not but that program got shut down when I stopped looking for stats/info.

The HTO Claro drivers are frustrating to get. Those should be updated per that same link you shared above though I will certainly double check. I will also check the box to make sure I have just the "Claro" and not the Plus or II version (which I'm pretty sure I do not).

Lastly, a couple follow-ups: Speaking with a tech guy I work with and since almost all the BSODs have been while playing Diablo III, he encouraged me to get a program that monitors heat (which I will tonight) but to also adjust the in-game option settings so that my Max Foreground FPS is capped at 60 max. He said if it's not heat and I still crash, to try setting the cap to 30 FPS and keep testing. Do you think FPS in such fashion could cause these crashes?

Finally, the memtest86 and memtest86+ tests where I only got errors during the block move test segments, do you think I should be concerned about my RAM being bad? They were just purchased in December. As well, G-Skill lists qualified MBs on their site and mine is not listed (G.SKILL-Products). I suspect it's because this is older RAM and that my MB was not available for verification/testing at the time. Do you think this is a concern bearing in mind when I had my GTX 570, all the things I normally do ran just fine with all the same hardware.

Actually, the crashes indicate graphics and network issues. They can be related to the drivers being loaded for your software that interfaces with the hardware with the Gigabyte utilities. Even though the software itself is not starting, the drivers are still loaded and may be causing conflicts.

Capping the FPS at 30 will put less stress on the graphics card, so it may help. It is more of a workaround than a solution, though.

Any errors in memtest86+ means you should probably RMA the RAM. Before doing so, though, you need to go through Part 3 of the tutorial to determine whether it is indeed a bad module and not a bad motherboard slot.

It could also be that your RAM and motherboard are not compatible and causing errors/crashes.

Thank you again, writhziden. Okay, I've just finished 3 hours of troubleshooting by testing my RAM and RAM slots on the MB. Believe it or not, all 4 chips passed and all 4 slots passed. Using memtest86+, I tested chips 1 and 2 in slots 1 and 2, no errors at all. I then tested chip 3 in slot 1 and it passed. Chip 4 in slot 1 and it passed. I then took chips 3 and 4 and test them in slots 3 and 4 ... passed. At this point I thought maybe I hadn't seated one RAM stick correctly or it had come loose. Placed all 4 sticks back in in all 4 slots and I got block errors on test #5.

My friend and I found a thread on G-Skill's site that stated we should adjust some BIOS settings (CAS and a couple others) so we did and I ran another memtest86+. I got 8 block move errors during the first 2 gigs to be tested, then 2 more errors on the final 16-17 gig test. I couldn't think of what else to do so I took some pictures with my phone so you can see my BIOS settings and the error codes from the memtest.

At this point, I'm thinking I may just need to tweak some more BIOS settings and I should be fine (at least I know I can run 8 gigs with 2 sticks and not have issues). Other than that, my only other thought is this RAM is not compatible with my MB.

An interesting case, for sure, and I super appreciate the help. I hope you're able to again!

Download and install CPU-Z and Upload screenshots of the CPU, Mainboard, Memory, and SPD tabs. In the SPD tab, upload an image of each slot.

Also, go into your BIOS and post all Voltages (CPU, RAM, NB, IMC, etc.) and all RAM settings (timings, frequency, etc.)

Thanks to Dave76 for help learning RAM Stability

Here's what I have. CPU-Z shots and since I'm not the most seasoned BIOS guy, I nabbed shots of most of the settings in there that seemed to be related to voltage (hope that's okay, wasn't sure what I was trying to find).

Your RAM is overclocked. Load setup/optimized defaults through the BIOS Exit screen, save, and exit. AMD processors do not like RAM running at 800/1600 levels without some major BIOS tweaks.

If the BIOS reset does not do the trick, I will walk you through steps to manually set the frequency.

Check the CPU-Z Memory tab again to make sure the DRAM Frequency is below 800 MHz after loading the BIOS settings.

Nevermind, I see your RAM is XMP. XMP works with Intel boards. It does not play well with AMD hardware. Try to find some non-XMP RAM for the system and see if the system is more stable.

Again, many thanks, writhziden. I'll go ahead and mark this as resolved. Fortunately, I ran into some contacts I have at NVIDIA who are offering to help me get everything sorted or new equipment if needed. So far, I have a couple techs claiming I can tweak a few more things but hey, if that doesn't work, RAM is fairly cheap so I am relieved either way.

Your help has been super educational and I'm now armed with more knowledge. For that, I can't thank you enough. Keep up the great work, I hope you know how much it's appreciated.

You're quite welcome. Glad you learned something in the process. Happy to see the thread solved.

Hope your friends are as helpful as they sound.