Computer crash and video goes blank.

Dooflop

Honorable
Apr 5, 2013
24
0
10,510
So I have an issue where my computer will randomly (usually when in a game) freeze for a brief moment, then turn off my monitors as if there is no video feed being sent to them...I can still hear people on teamspeak of mumble or whatever VOIP im using at the time, but i cannot talk back to them. So my computer is not receiving any input. Doing a hard restart fixes it just fine, in my event viewer log there is nothing video driver related. I have the logs from system events and administrative events to post here if need be. Im posting in the motherboard forums because that is my current suspect. I have run a memory test which came back negative for any issues. My video card is relatively new so I don't believe it to be the issue but I have not ruled it out. My PSU would be my second guess after motherboard...If more information would help in this scenario please ask and I will get that information.

My current setup:
Windows 10 professional
Asus sabertooth x79 motherboard
Intel Core i7 4930k processor
Nvidia Titan X(pascal) GPU
Corsair HX850w PSU
G,SKILL 16GB @1600MHZ 4x4GB
Intel 530SSD 480GB (boot drive)
Intel 330SSD 120GB
WD Green 1TB HDD
 

Dooflop

Honorable
Apr 5, 2013
24
0
10,510
Pretty good temps, highest ive seen is 76c and thats when it's working its ass off.
Also i can't quite remember when the issue started, its been a while...but it did never happen with my old GPU. I used to get the same crash before and when i went to restart my computer it would say "Overclocking failed" so i turned off my XMP memory profile and now instead of getting an overclocking failed screen on reboot it just reboots normally. Also I ran the XMP profile for quite a long time and its never given me that message until the new gpu.
 

genthug

Honorable
So, at this point you aren't running the XMP profile. What happens when you run a memory tester like memtest86? Is your CPU stable through a test of Prime95 or an equivalent CPU stress tester? Is your GPU stable through tests of Firestrike, through tests of Unigine Valley, Heaven, etc? If you haven't run all of these tests yet, I wouldn't be so quick to jump to the conclusion of the motherboard or PSU, as they are both nice pieces of hardware.
 

Dooflop

Honorable
Apr 5, 2013
24
0
10,510
Submitting a followup, I ran memtest86 for about 4 and a half hours for 2 passes and got 0 errors returned. Also last night when i was watching netflix my computer suddenly decided to stop the show, I then noticed I had 2 messages in my windows 10 message center, i clicked on it and nothing happened. Then i tried to reopen mozilla firefox to restart netflix, it would not open. I then realized at this time my entire computer (while not being frozen) was unresponsive. I could not open any programs at all, after a restart everything was fine, but i noticed i had a message in my message center that said "are you having sound problems?" I am going to run some GPU intensive tests now.

Edit: Just ran 3Dmark Timespy, Firestrike, and Firestrike ultra stress tests on my gpu. All of them took around 50 minutes to complete. Highest temperature achieved was 73c and I had no artifacting or crashes. Should I run these longer or move on to prime95 testing? Also what would be the best settings for prime95 testing and for how long should i test it?
 

genthug

Honorable
I would move on to prime95. Small FFTs, one hour should suffice... If you're not overclocking at all I wouldn't expect it to crash. Which would return us to your initial suspects, but for clarity's sake, might as well run P95. Just to make sure.
 

Dooflop

Honorable
Apr 5, 2013
24
0
10,510
well it crashed again today... swapping cables on my PSU did not fix the issue. Maybe my PCI-E port is broken on my mobo and not feeding it power properly? Should my next step be to swap it into my x8 slot? And will that effect the performance of a titan x pascal?
 

Dooflop

Honorable
Apr 5, 2013
24
0
10,510
I had two GTX 770 2GB SC's in SLI, Gave one away to a friend but I do still have one.

just swapped my gpu into a different PCI-E slot...Which turns out to actually be another x16 so i shouldn't see any performance loss. I will re-post with results, good or bad. Want to help people avoid this issue as much as possible, its awful.
 

Dooflop

Honorable
Apr 5, 2013
24
0
10,510
Alright, just crashed again in a new PCI-E slot, still using different cables then originally. Running out of options here...I dont know if testing my other GPU (gtx 770) will do me any good because it's not as power demanding as the titan x pascal. What should I do now? I don't have any other power supplies to test with.

Just wanted to mention, I wasn't doing anything GPU intensive when it crashed today. Was literally just reading an article online.
 

genthug

Honorable
It will give a base line. Your system was running fine without the Titan XP in it, and then you put the new card in, and it's suddenly crashing. Putting the old GPU back in would be a good place to start with as it would allow you to test if it was the GPU causing the crash within your machine (i.e. we can rule your motherboard out).
 

Dooflop

Honorable
Apr 5, 2013
24
0
10,510
SO my computer just lost power and restarted itself. like 95% sure its PSU now. probably going to buy a new one.
if anyone has some last suggestions before i blow money, please advise.

also any PSU suggestions? im thinking a higher end corsair.
 

Dooflop

Honorable
Apr 5, 2013
24
0
10,510
sooooo it crashed after i replaced the PSU, this got me to messing around with the mobo and bios, i downloaded the asus ai suite which alerted me that my VCCSA temperature was 124 degrees celsius. I'm not sure if this is accurate or if my asus ai suite is incorrect because it ALWAYS reports 124.0 degrees celsius. How can i know if its giving me false temps?...if its not I'm no wizard but i think my Mobo is broken. Am i correct in that assumption? My motherboard is covered by warranty so...there is that. Is there any way to check if its giving me false temperatures though? anything else i can do before i warranty this and it just ends up being a bad thermometer? I'd hate to go a couple weeks with no PC just because an incorrect temp reading.

(good news is i can return the PSU)
 

genthug

Honorable
If it's covered under warranty, I'd RMA it then. That's... literally the only other piece of equipment in your machine that hasn't been checked. I would return the PSU but at this point--your CPU has been stressed test, you've run memtest, you've run Firestrike, Valley, etc. and they've all passed, you're not overclocking, your PSU isn't an issue as you've replaced the entire thing and you replaced the old 12V rails... you have one piece left, and that's your motherboard. I've never liked AI suite for it's temp monitoring, I've always used HWMonitor. AI Suite gives temperatures and relates them to things that they... aren't. That being said, their temperatures have always been accurate even if their labels are not. Your VCCSA temps should only be high if you're overclocking, which you're not.
 

Dooflop

Honorable
Apr 5, 2013
24
0
10,510
SO im posting back as I have solved the issue and would like to spread the news that this should be done if this happens to you. After all the stress tests and hardware replacements etc...it turns out all I had to do was update my chipset drivers and I flashed my BIOS, now im not sure which of the 2 fixed it, as i did them both at the same time. But I have not had a crash in 9 days now, using my old PSU and my graphics card that had the issue. So to anyone with this issue...try updating your BIOS and chipset first.