Can someone help me interpret my GPU sensor log and troubleshoot ?

[00zim00]

Hey, I have been dealing with a driver crash issue "Display driver stopped responding" then BOD for a few months now. One person asked for me to make a active log using TechPowerUp GPU-Z Sensor program which I have done.

LOG LINK: https://www.dropbox.com/s/eauzwpo19zjvbag/GPU-Z Sensor Log.zip?dl=0
The crash happen around 716pm, which you can kinda see at the bottom of the log.. i think

I have been avoiding playing any games which require excess memory usesage to avoid haveing an issue as i didnt have the time to fix it if something permanently broke my pc.. however.

Yesterday I tested and played Civ5 for an hour or so without an issue. Today I was playing this game again and all of a sudden after about an hour my computer started getting choppy and freezy, then a taskbar pop-up appeared saying my display driver had crashed. I attempted to close everything that was open in order to get as much data as possible but after 10-15 secs my pc BOD with the error. DRIVER_IRQL_NOT_LESS_OR_EQUAL.

Previous photos I was able to take: http://imgur.com/a/lJR46
Wasn't able to do the same this time but they are basically the same.
Original post: https://www.reddit.com/r/AMDHelp/comments/4iqnz3/how_to_test_if_my_card_is_broken_or_if_its/

I then was unable to restart my pc without it saying it had to instantly reset. After about 10 times trying I was able to get in twice where it would have both my screens showing the graphical artefacts at which point 5secs later my pc would just crash again. After about 15 mins of trying i was able to get in my pc again but pressing ESC a few times durin start up, I was attempting to start in safe made however that didnt work yet miraculously when my computer started up after that fail.. everything looks ok again and is running... as long as im not useing anything graphics intensive that is.

It dosnt appear to be heat related as first suspected by people, so beyond that im stumped and was hoping someone had an idea based on the log files.

Computer Type: Desktop

GPU: AMD Radeon R9 200. Have not over clocked anything.

CPU: i5-4670k 3.4

RAM: 16 GB 'G.Skill Sniper F3-12800CL9D-8GBSR 8GB (2x4GB) DDR3' x 2 = 16gb

PSU: ToughpowerXT 875w TPX-875M AU

Operating System. Windows 8.1 Pro

GPU Drivers: 16.15.2211

Chipset Drivers: NA

Background Applications: Random

Description of Problem: Driver Crashing, screen flickering, pc restarting itself/crashing.

Troubleshooting: Updated drivers, Reset pc, made sure card was in correctly.

 

[W1zzard]

Clock down both GPU and memory by, say, 50 MHz, maybe the card isn't 100% stable at stock clocks.

Also try the card in another PC, or try another card in your PC - to isolate whether it's a problem with the card or your PC.

 

[00zim00]

Clock down both GPU and memory by, say, 50 MHz, maybe the card isn't 100% stable at stock clocks.

Also try the card in another PC, or try another card in your PC - to isolate whether it's a problem with the card or your PC.

Im kinda not really tech savy.. so are you saying from the logs there is a peak in MHz or something before the crash so thats why i should lower it?
Also if i did cloak it down is that something i would do just though the AMD settings app?

Lastly just for a point of reference, the card didnt have any issues for about 8 months, then it randomly started like 6 months ago.. If that makes a difference.

 

[W1zzard]

I didn't even look at the log, I don't think it has any useful info.

Yeah in Radeon Settings (right click on desktop), then Global and Global overclocking or something.

 

[DeathtoGnomes]

https://www.techpowerup.com/forums/account/specs

Fill that out and if you can take the side off your case and post a couple pics.

Do remember any events when this started to happen, like installing a program or updating windows, etc.

IF, if, this is from corrupted video drivers, start with downloading the latest driver for your card and also the AMD Cleanup Utility (If that doesnt work try DDU here. )Run the cleaner first then assuming no problem there, reinstall new drivers.

If this is from anything else, you may have to update your motherboard drivers, which you havent listed yet, or you may have to reinstall windows.
.

 

[Naki]

GPU: AMD Radeon R9 200.

This is too vague. Please provide exact video card brand & model. Such as GIGABYTE, ASUS, MSI, SAPPHIRE, XFX/etc. And exact model such as R9 280, 280X, 290, 290X/etc.
If you cannot answer this, just give us a GPU-Z screenshot with your card's info.
Also, what motherboard do you use? You can check it via CPU-Z or Speccy or AIDA64.

I think either - your video card is damaged, and/or RAM and/or motherboard. PSU might be an issue too.
Did you check your RAM to be working or faulty? You can do so via MemTest86:
http://www.memtest86.com/

Your PC has a hardware issue. Playing or not playing any games will not magically fix a hardware issue in any way. You need to find out which component(s) is/are faulty/damaged, and replace it or them.
IF the PSU is damaged, it could damage the videocard, and motherboard along with it.

I suggest you find someone that can properly diagnose the issue. Such as local computer repair shop/etc.

 

[Naki]

Hello? Is the issue fixed, please?

 

[00zim00]

@Naki @DeathtoGnomes
Hey! Sorry im late to reply. I starting writing a response that got very long. I then stopped half way though as it was late then I forgot to reply!

Fill that out..specs:
Done, also here is a screen of GPU-Z

Downloading the latest driver for your card and also the AMD Cleanup Utility:
The AMD Cleanup Utility link within that page dosnt work any more saying its a incomplete location. So when I get time next ill try DDU as suggested. As for the MB update I could do that or reinstall windows, I was hoping to avoid having to reinstall windows but if worse comes to worse I might have to do that anyway. Ill keep that as one of my last resort options

Do remember any events..:
Ill give as many details as possible, im not sure what is helpful and what is not so i will write everything I remember of note..
Also im sorry if its way to long or confusing, im not the best at explaining these things which I know can be frustrating

First Main Crash:
The only event of note I can remember happening is that it happen a few weeks before Xmas of 2015, It first happen after I had been playing XCOM: Enemy Unknown for a few hours after playing it most days within that week. It was what appeared to be using way to many resources and heating up my pc. However I was in summer during that time so was hard to tell plus Xcom had been known to be a resource hog. The first crash happen with that game. I tried to troubleshoot the game itself at the time thinking it was just Xcom not liking my PC, I had found people having similar issues in the past so at the time wrote it off to just that game.

Later on I tried to manage what I thought was a heat problem after cleaning my pc and checking connections, installing an extra fan and moving my PC case off the ground to a more ventilated position. I thought over heating wasnt the issue just a coincidence but at the time I couldn't be 100% sure. However after keeping an eye on it for a few months aswell as using programs to monitor my GPU/CPU heat I felt more confident that heat wasnt the issue. Likewise recently when the last crash happen, the heat of my pc at the time was only about 62D and I had only been playing Civ 5 for about 45mins. Where the day before i had played Civ 5 for a few hours with no issue.

looking back again, another thing I remember is that not far into January reports were coming out about a faulty AMD driver which caused your fan not to speed up causing overheating. This in my mind explained the original overheating but since I had no other crashes at that time I couldn't confirm either way. Likewise by now there also have been driver updates that supposedly fixed that fan/heat issue. An seeing as I still randomly get the crash and while using GPU-Z I have found that heat dosnt seem to be an issue as the heat as said before is relatively low most of the time never hitting the critical 80+D it had gotten with that driver/xcom game.

Second Main Crash:
I was playing stellaris (release day May 9, 2016), I had been playing for about 8h when my monitors started randomly getting those artefacts, I closed the game and noticed the Display driver had stopped with the pop-up coming up "Display driver stopped responding and has recovered". After reseting my pc a few times i was able to start to pc without the graphics artifacts from showing up. It seems that until the pc is reset a few times an "fixes" itself It can be prone to crash for a few hours to a day or so until randomly its back to normal. Im guessing something in memory isnt cleared until after some time which removes the offending cache data.

Third Main Crash:
Another similar occurrence was with the game Prison Architect, which is a very low resource non intense game. The "Display driver stopped responding and has recovered" appeared while playing not long into the game. This case however I had noticed windows 10 had decided to download itself in the background, as well as steam doing a few GB updates which made my C drive only have about 600mb left down from about 10-15gb free. So im guessing this case was unrelated.

Recent Main Crash:
Whats in the first post. Happen after playing Civ 5. The BOD error messages seem to be random and never the same one, from what I can tell. Likewise I havnt had any more crashes since then, but I havnt been playing the game, I plan to play again this weekend. One thing to note aswell is that not all games make my pc crash, I can play for example MTG Online though steam or play the Binding of Isaac and have no issues at all.


I suggest you find someone that can properly diagnose the issue:
I did try that after the first crash but all it did was cost me money and all they did was make sure my video card was connected correctly lol. But your right i should look around and try to find a computer shop who is competent, just isnt many around here.


Did you check your RAM to be working or faulty?
This weekend I plan to try memtest as well as reinstall clean drivers to see if anything else happens when playing a game. The difficult part is I can have alot of time with no issues which makes me think what I did fixed it and its working again only to have it 'randomly' happen again lol.



Thanks again! If you feel its better for me to just to more tests before you reply thats fine with me The information i have given probably isnt that helpful.

 

[mstenholm]

C'ome on AMD drivers have given that errors in all the years I have been folding on AMD GPUs regardless of drivers and GPUs (7770 up to 280X). The solution is, as W1warrd suggested to down-clock. I can see that you don't OC but still try the 50 MHz lower clock.

 

[Naki]

No. I have 3 AMD Radeon cards here, and never have such issues. Namely HD 7770, R7 260X, R9 290. Not the issue here. User is also not doing any folding, so mentioning that does not help any.

EDIT: Additionally, I occasionally fold on my R9 290 card & Core i7 CPU - this is never any issue (except the load it puts on CPU+GPU).
EDIT #2: I did get some Folding@Home issues similar to what you report/mention with older AMD cards, such as HD5650M and HD4870 before that. Those are fixable via replacing some Folding@Home system DLLs with other versions. Info in Folding@Home forums.

The OP's mobo/platform is quite/fairly old. Maybe mobo has decided to go where all older motherboards sooner or later go - the "happy motherboard hunting grounds".

 

[Caring1]

If it was artifacting after or prior to the crashes I would suspect the Vram on the card, especially Elpida Ram.
Have you tried another PSU just to rule out the power hasn't degraded to the GPU?

 

[DeathtoGnomes]

Awesome crash report @00zim00 ! Yea overheating may have started everything down this road. I would bet the thermal pad/paste on the GPU is not conducting heat very well. See the problem with heat is that if you have poor cooling, over time, everything will just become weaker, even the current cooling setup can deteriorate on you, creating yet even more cooling issues.

If you feel up to the challenge, buy some decent thermal paste and some thermal cleaner solvent and re-apply on your GPU.

If not, then under clock your GPU as suggested, and when gaming, take the side off the case and put a room fan next to the case and point it at your video card.

 

[Naki]

Really bad idea. I am unsure if OR has ever done this operation you mention - if not, this is a bad time to try this now first time.
We do not know if OP has 2 (3?) left hands here, so a really bad suggestion.
Some of us do - we can replace easy stuff like RAM or HDD/SSD or video card, but cannot perform complex/delicate operations, and thus if such are necessary, we usually ask for help from a skilled friend, or carry our PCs to a local PC repair shop for this.
Not all people are handy with handling PC parts, please note.

Also, issue might not be the videocard at all. Could be motherboard issue, we cannot rule that out at this point. PSU might be an issue too.

EDIT: OP, thanks for GPU-Z screenshot! This is a GIGABYTE AMD Radeon R9 290 card - checked Wikipedia. I have a very similar card, same specs, but mine is SAPPHIRE and has Hynix RAM instead of Elpida.

 

[DeathtoGnomes]

We all have to get our feet wet at some time, have some confidence in the OP, he has options, he can attempt it or he can get someone else to do it. Ofcourse, direct supervision for that would be a big help the first time, but sometimes you just gotta deal with it and jump in the water. The worst that can happen is he has to get another video card to replace a card thats already failing.

And if you got some solid input on what IS wrong, stop with the shoulda, woulda, coulda's and contribute something other than "might be bad".

 

[Naki]

It is difficult to diagnose a hardware issue remotely, especially when person with said issue not having much experience with such issues.

I think if OP can borrow a known to be working video card from a friend, that would help.
Card should be similar model, such as maybe AMD R9 270X or higher.
IF a working card is failing, then the computer is damaged - mobo, and/or PSU/etc.
IF the working card works, that will mean the video card - AMD Radeon R9 290 in this case - needs replacement and hopefully mobo & other parts are still fine and undamaged.

Trying another PSU is also one of the 1st things I would try, if I had these issues.

 

[00zim00]

(Is there a way to hide/shrink the images so they dont take up so much space? Im sure its hard to read otherwise :O!)

Im still in the process of troubleshooting but thought i would give an update since there have been so many helpful responses!

C'ome on AMD drivers have given that errors in all the years I have been folding on AMD GPUs regardless of drivers and GPUs (7770 up to 280X). The solution is, as W1warrd suggested to down-clock. I can see that you don't OC but still try the 50 MHz lower clock:
Just to make sure before I do whats suggested. Is this the setting im supposed to change?

The OP's mobo/platform is quite/fairly old.:
Its a new motherboard, well about 2 years old now.. i know the model wasnt the best but at the time I just got a motherboard within my budget that worked with the parts i wanted to get. If its the MB that pretty much means i might as well get a brand new pc seeing as getting something thats compatible with the parts I have seems kinda difficult compared to just getting everything new :/.. then again im terrible at understanding what hardware works with what lol

Start with downloading the latest driver for your card and also the AMD Cleanup Utility:
(Edit) I asked on the AMD support forums to make sure it wasnt just a broken link and they also just suggested useing the DDU Uninstaller so I will use that next time)

I tried to install the AMD driver as a "clean install" but I have a feeling that the driver wasnt uninstalled first. I just used the inbuilt AMD driver repair/reinstall as the AMD ultility dosnt work on the site.. Ill have to do it again correctly however i did try to play a game after I did this.

I stared Civ 5, played for less then 10 mins, pressed next turn and.. In game error and driver crash, artefacts, ect.. only made it to turn 3 :/.
GPU-Z Log:
https://www.dropbox.com/s/ungjycad8j3ubti/GPU-Z Sensor Log2.zip?dl=0
Game Error:

Is there a log somewhere that keeps a record of WHY the driver actually stop working? Or is that information not actually kept anywhere. I was thinking about installing an old driver from 10/20/2015. Would it be worth installing an older driver to rule out the newer drives breaking things?


I think if OP can borrow a known to be working video card from a friend, that would help:
My Brother has a near identical PC to mine, in fact he has the exact same graphics card as we brought them at the same time at the same place.
From memory he has the same CPU (tho he has a CPU cooler), same graphics card but different Ram, Motherboard and PSU.
I could probably convince him to borrow his GPU but he would murder me if my PC breaks it on him lol. I dont think he will let me use any other parts, especially if I would have to pay to replace them if they do break :/. Although maybe thats a risk I need to take. On the plus side im still in warranty for all/most of the parts in my pc i think.

Trying another PSU is also one of the 1st things I would try, if I had these issues. // Have you tried another PSU just to rule out the power hasn't degraded to the GPU?
Im curious how the PSU would be the issue?(Noob here) Wouldnt my PC turn off and not just have the driver crash, artefacts and then sometimes BOD?

If you feel up to the challenge, buy some decent thermal paste and some thermal cleaner solvent and re-apply on your GPU. //Really bad idea. I am unsure if OR has ever done this operation you mention - if not, this is a bad time to try this now first time:
I have to admit im not confident enough to do that myself, if I work out that its the GPU thats at fault I will most likly get it replaced under warrently, plus wouldnt the warranty be void if i started doing things to the GPU?


Some more random info that came to mind that might be helpful.
-I do have a portaable HDD (just used for storage) connected to my pc, the games are not running from it but I didnt think of mentioning it before has I have no reason to believe it could be the problem but thought I would mention it now.
-I have set the TdrDelay to 8 which I had heard could help, hasnt done anything.
-I still plan to do the memtest and 3d bench mark, I just need to make sure I have time to monitor it when im doing the tests which hopefully I can do soon.
-The motherboard driver updates are done in the bios menu when you start up your pc right? or is it like the GPU drivers where I just install something on my desktop?
-Im not sure if i missed any other suggestions.. I am rereading suggestions so I will get around to all of them as i slowly rule things out.

Thanks for the help so far!

 

[Naki]

You are welcome.
Usually, you do not need to monitor the Memory Test much.
In fact, usually it is a good idea to let it run at least 2-3 passes, during which time you can be away from PC (sleeping/at work/etc).
However, as it may restart or crash in your case, maybe it is a good idea to monitor at least the first pass.
2-3 passes should not take much time if you boot the USB flash drive with Memtest86 on it in UEFI mode.
Please note for best results you should boot the USB drive that way - if you boot into Legacy mode, an older and slower version of MemTest86 will run.
(the drive will have TWO separate Memtest86 versions on it once you create it)

 

[mstenholm]

(Is there a way to hide/shrink the images so they dont take up so much space? Im sure its hard to read otherwise :O!)

Im still in the process of troubleshooting but thought i would give an update since there have been so many helpful responses!

C'ome on AMD drivers have given that errors in all the years I have been folding on AMD GPUs regardless of drivers and GPUs (7770 up to 280X). The solution is, as W1warrd suggested to down-clock. I can see that you don't OC but still try the 50 MHz lower clock:
Just to make sure before I do whats suggested. Is this the setting im supposed to change?

Yes and no. First step would be to down-clock the GPU. Your RAM (memory) which is what you are about to down-clock is generally not a problem for the 1500 MHz cards. AMD partners did a one point ship the GPUs with 1600 MHz clock and in these cases a down-clock to 1500 could save the day.

 

[00zim00]

@Naki, I will keep that in mind

Yes and no. First step would be to down-clock the GPU. Your RAM (memory) which is what you are about to down-clock is generally not a problem for the 1500 MHz cards. AMD partners did a one point ship the GPUs with 1600 MHz clock and in these cases a down-clock to 1500 could save the day.

Oh I see, so the GPU Clock which is default at +58.3% put it down to like 50% or until the GPU MHz which is now 399MHz is showing 349. (If im reading the program correctly)

 

[Naki]

I bet under clocking the GPU will not help any. Willing to bet real money too, if you want.

 

[00zim00]

I bet under clocking the GPU will not help any. Willing to bet real money too, if you want.

So a bet between you and mstenholm? loser buys me a new GPU or PC !

 

[mstenholm]

So a bet between you and mstenholm? loser buys me a new GPU or PC !

Consider yourself lucky that we spend this much time to help you....over and out

 

[00zim00]

Consider yourself lucky that we spend this much time to help you....over and out

And I do appreciate it I was just joking.

 

[PainfulByte]

Just checked your initial log, there is a spike in the VDDC output (more than double the usual output) @ 19:16 after the core having maintained its highest logged clock(947Mhz for 5 or 6 seconds, which is also higher than the rest of the logged speeds). This happens at the same time the memory clock drops down to its minimum speed. So this looks like the crash moment.

I agree with Wizzard, the logical step one would be to downclock core/mem by 50Mhz.

A good night-long memtest run would be also a good step.

Beyond this, everything else is guessing at this point.

 

[DeathtoGnomes]

I didnt know the card was still under warranty, so yes, re-pasting the GPU would void its warranty. (sticker over a screw? LOL) Better to takes steps to RMA first, should the problem be the card itself and not something else.

Since thats the case, you should contact customer service and follow their troubleshooting steps so you can RMA the card, because sticking it in another system might also void warranty. However, trying a different card in your system will help troubleshoot what is left in your system.