Black Screen + Reboot During Gaming (No BSOD/Error Code)

Baasha

Golden Member
Jan 4, 2010
1,989
20
81
For the past few weeks, I had been having issues with my X79 rig where I'd get black screen and reboots only during gaming - Batman Arkham Origins, BF4, AC4, etc. etc.

At first, it seemed like a power issue so I got a new PSU - Lepa G1600.

Still kept happening - tested each GPU individually and all passed without error/problems.

Black screen problem still persisted. So I installed Win 7 x64 on a spare HDD and checked to see if it was software/drivers - still got black screen/reboot.

Currently on Win 8.1 Pro x64 btw.

Decided it had to be the motherboard - RMA'd the board and just got a replacement (2 days ago). Still get black screen/reboots.

Reflashed the GPU BIOS - worked for a while when the voltage was left at stock (instead of 1.212V). However, still got black screen/reboot after about an hour. Then tried NFS Rivals and got BSOD (0x19) four times.

0x19 pointed to "BAD_POOL_HEADER" so I decided to test RAM w/ MemTest86+.

Tested RAM @ XMP setting (Corsair Dominator Platinum 32GB 9-11-11-31 @ 1.65V) - got a ton of errors in MemTest86+ (411 errors after about 8.5 hours).

Reset RAM to "Auto" (1333Mhz) and booted into Windows. Tried playing AC4 and got black screen + reboot within 10 minutes.

I am at a complete loss as to what could be causing this issue.

I had an electrician install a 20A breaker in my place just for the rig - have a heavy duty 12AWG cable for the PSU.

The one thing that seems to be "weird" is that I cannot for the life of me update the chipset drivers on my mobo.

The version I have now is 9.2.3.1032 while the one on ASRock's website is 9.4.0.1026. There is an updated version 9.4.4.1006 but every time I run the installer (either through EXE or through Command Prompt), the window pops up and immediately says "successful". When I check Device Manager, the version is still the same 9.2.3.1032 and has NOT been updated.

Pic:
feeaz4.jpg


Could this be the culprit?

Other than that, I have NO idea why this is happening or what's causing it.

I also tried a different socket (15A) just for testing and got the same result - black screen & reboot.

Would really appreciate help with this.

Details on the MemTest I did: http://forum.corsair.com/v3/showthread.php?t=124397
 

Virgorising

Diamond Member
Apr 9, 2013
4,470
0
0
Wow. Talk about dense troubleshooting! Impressive.

Just for the heck of it, try downloading yr chipset drivers directly from Intel and see if it works and they install:

Download Center
 

Ketchup

Elite Member
Sep 1, 2002
14,553
248
106
Couple things:

If you play the game with one video card, do you see the issues?

Just because you have lowered the memory speed just mean you are automatically making the errors go away. If not one of the video cards, sounds like you have a bad stick of RAM.
 

Baasha

Golden Member
Jan 4, 2010
1,989
20
81
How's the memtest look at the Auto settings?

The Auto settings (ran only 1 pass) seemed to be fine but I still got black screens during gaming.

I ran the RAM individually today (1 stick at a time) - at the XMP setting (2133Mhz).

Three of the sticks passed fine but the last one (#4) gave 10 errors in the 1st and 2nd pass.

I took two of the sticks out - put the two "good sticks" in, set the XMP setting and booted into Windows.

Ran benchmarks - passed great even @ 1.212V @ 1150Mhz (my old OC). Thought that this was "it".

Fired up AC4 - black screen within 5 minutes. :(

Tried BF4 - DirectX Error - game won't even start (is it because 16GB system RAM isn't enough?).

Tried BF3 - crashed within two minutes - black screen & reboot.

I literally have NO freaking clue as to why this is happening.

The GPUs work individually - they even passed the benchmarks together - obviously, something is causing the GPUs to not work well when gaming.

I will try another driver reinstall but I highly doubt that will have any positive effect. :thumbsdown:
 

Baasha

Golden Member
Jan 4, 2010
1,989
20
81
Couple things:

If you play the game with one video card, do you see the issues?

Just because you have lowered the memory speed just mean you are automatically making the errors go away. If not one of the video cards, sounds like you have a bad stick of RAM.

Well, the bad stick of RAM has been confirmed - #4 gave errors in the first and second passes while the other 3 sticks worked great at their rated 2133Mhz @ 1.65V.

The GPUs work fine individually - I am able to OC them to 1150Mhz @ 1.212V. Three of them are stable at 1189Mhz and the 4th is 1150Mhz.

After installing just 2 sticks of "good" RAM, I ran the benchmarks @ 1150Mhz (Heaven 4.0 & Valley) @ 1.212V and they passed without a single hitch - ran them for several loops each.

The moment I start playing a game - AC4/BF4/BF3/NFS Rivals etc. - I get a black screen & reboot within a few minutes. Sometimes it takes 20 minutes, sometimes 2. Is there a 'cry' emoticon? :cry:
 

Matt1970

Lifer
Mar 19, 2007
12,320
3
0
I have seen device manager display the wrong driver version before. Re-run the chipset driver update and see if it actually tells you what version it is updating from while it installs.
 

Baasha

Golden Member
Jan 4, 2010
1,989
20
81
I have seen device manager display the wrong driver version before. Re-run the chipset driver update and see if it actually tells you what version it is updating from while it installs.

Yea, that's a good point. I checked that and it does show 9.4.4.1006 while installing - so it looks like it did update. Also confirmed it with someone else with the same mobo on another forum - he too shows the 'old' driver in Device Manager.

I don't think that that's the issue though. :(
 

richaron

Golden Member
Mar 27, 2012
1,357
329
136
Well, the bad stick of RAM has been confirmed - #4 gave errors in the first and second passes while the other 3 sticks worked great at their rated 2133Mhz @ 1.65V.

Yes, you are describing the symptoms of unstable RAM. And you have identified a "bad" ram module.

Diagnosis is complicated by OC/UC & OV/UVing (or I missed something). Is there a difference between good and bad pairs of ram with everything else stock.
 

Ketchup

Elite Member
Sep 1, 2002
14,553
248
106
Well, the bad stick of RAM has been confirmed - #4 gave errors in the first and second passes while the other 3 sticks worked great at their rated 2133Mhz @ 1.65V.

The GPUs work fine individually - I am able to OC them to 1150Mhz @ 1.212V. Three of them are stable at 1189Mhz and the 4th is 1150Mhz.

After installing just 2 sticks of "good" RAM, I ran the benchmarks @ 1150Mhz (Heaven 4.0 & Valley) @ 1.212V and they passed without a single hitch - ran them for several loops each.

The moment I start playing a game - AC4/BF4/BF3/NFS Rivals etc. - I get a black screen & reboot within a few minutes. Sometimes it takes 20 minutes, sometimes 2. Is there a 'cry' emoticon? :cry:

OK, I think we are narrowing things down. Is there anything in the minidump folder from the crashes without the bad memory sticks installed?

Also, if you run a single card from the second pci-e slot, can you replicate the issue?
 

Mushkins

Golden Member
Feb 11, 2013
1,631
0
0
When you get a black screen crash, does the audio sort of stick/loop for a little bit?

This is typically indicative of what happens when the video drivers crash and cannot recover. I used to get these with Skyrim (and only Skryim) on my old 5770, tried a billion different drivers, patches, and updates. The only solution was to buy a new card.

Sounds like one of your cards might be slightly defective, causing the whole thing to fail selectively when running SLI/Xfire. Try some extensive gaming with different configurations of cards and see if you can narrow it down to a specific card causing the issue.
 

Baasha

Golden Member
Jan 4, 2010
1,989
20
81
Sounds like one of your cards might be slightly defective, causing the whole thing to fail selectively when running SLI/Xfire.

This was EXACTLY the issue!

It turned out that the 3rd GPU was not able to hold 1.212V.

I tried 3-Way SLI testing and that GPU failed (system crash).

I put the 4th card in for another 3-Way testing and it was able to not only hold 1.212V, but keep my old "OC" perfectly well! :D

I was able to do benchmarks, play games, and even render a sample video without issue!

Time to RMA that GPU!
 

Mushkins

Golden Member
Feb 11, 2013
1,631
0
0
This was EXACTLY the issue!

It turned out that the 3rd GPU was not able to hold 1.212V.

I tried 3-Way SLI testing and that GPU failed (system crash).

I put the 4th card in for another 3-Way testing and it was able to not only hold 1.212V, but keep my old "OC" perfectly well! :D

I was able to do benchmarks, play games, and even render a sample video without issue!

Time to RMA that GPU!

Glad we worked it out :) It drove me nuts when it happened to me, but it *only* happened in Skyrim so I just ignored it until I built a new PC.
 

Baasha

Golden Member
Jan 4, 2010
1,989
20
81
Glad we worked it out :) It drove me nuts when it happened to me, but it *only* happened in Skyrim so I just ignored it until I built a new PC.

I think I spoke too soon! :mad:

It was fine until I fired up NFS Rivals today when the game just froze about 10 minutes into a race.

I was playing NFS Rivals everyday since I thought I had 'solved' it and there were no freezes/crashes/hangs.

I played AC4 for about an hour and same thing - freeze!

This time, however, the game just freezes (NFS Rivals had a 'ding' sound after the freeze) with the picture - no black screen and/or reboot. I had to hard reboot so I got the stupid Event ID 6008 ('unexpected shutdown') error but other than that, no error codes, BSODs, or BugCheck codes(?)!

I am seriously about to kick my computer!

I tried both 331.82 WHQL and 331.93 Beta drivers and they both do the same thing.

However, I also don't think it's a GPU-voltage issue like before since I am playing at my "regular" OC: 1.212V @ 1150Mhz and +300 on the Mem.

NFS Rivals froze within 15 minutes four times.

AC4 froze after an hour of gameplay.

I'm going to try BF4 now and see what happens.

Any advice?
 

Ketchup

Elite Member
Sep 1, 2002
14,553
248
106
... I was playing NFS Rivals everyday since I thought I had 'solved' it and there were no freezes/crashes/hangs.... This time, however, the game just freezes (NFS Rivals had a 'ding' sound after the freeze) with the picture - no black screen and/or reboot. I had to hard reboot so I got the stupid Event ID 6008 ('unexpected shutdown') error but other than that, no error codes, BSODs, or BugCheck codes(?)!..

I was having similar issues with NFS Rivals, but I could always Ctrl+Alt+Del out of them. Once I got the beta drivers (same as you) the issues just about stopped for me. Now I just get a DirectX error every once in a while, but I figure a game update will fix that.

Denis280, I think he is running a Lepa G1600, which should be more than enough.

Let us know how the other games do Baasha.

If you are still seeing issues, I would recommend trying to recreate the issues on a single card.
 

Baasha

Golden Member
Jan 4, 2010
1,989
20
81
I was having similar issues with NFS Rivals, but I could always Ctrl+Alt+Del out of them. Once I got the beta drivers (same as you) the issues just about stopped for me. Now I just get a DirectX error every once in a while, but I figure a game update will fix that.

Denis280, I think he is running a Lepa G1600, which should be more than enough.

Let us know how the other games do Baasha.

If you are still seeing issues, I would recommend trying to recreate the issues on a single card.

Hey ketchup,

Thanks for the follow up.

I was pulling my hair out over this issue but finally solved it!

It turned out that the Lepa G1600 was the culprit the whole time! :mad:

Whenever it was near or above 1500W at the wall, the computer would shut off.

So I tried both my old Antec HCP-1200 and the Lepa together and it worked fine. Still got freezing in some games though - BF4, AC4, NFS Rivals etc.

Then, I just tried the Antec by itself and guess what? No problems at all! :D

I ran many benchmarks and played BF4 maxed out for over 2 hours without issue! Other games work well - NFS Rivals still freezes but I think it's some sort of software/game bug(?).

Saw the Antec pull 1723W at the wall without skipping a beat! What a beast of a PSU! :)

I ran 3DMark 13 Fire Strike Extreme and took the top spot on OCN:

a77aab1e_2v9cjrq.jpeg


Returned the Lepa and now on the hunt for another 1500+ watt PSU.