newly bought GTX 970 crashing most games and has glitch

Toggle sidebar Toggle sidebar
I

isisisaac

Distinguished
Nov 1, 2008
22
0
18,520
HI guys,

lNBMLk5.jpg


Just got my Zotac GTX 970 2 weeks ago. it passed 3D mark benchmark few times without problems, but several games that I play they often crash once in a while randomly. It can be as short as every 5 minutes to once in 2 hours.

Sometimes after a while during game play, I see some glitches on the screen, restarting the game could solve this problem, but sometimes it just crash and return to windows. When the game crashes, most of the time I see this Desktop Notification saying "NVidia display driver stopped responding and has recovered"...., sometimes I dont have this notification and the game just shutdown itself.

Sometimes if I restart the PC, then it will not crash during gaming until say 1 or 2 hours. It's just random.

Games that i have problem.
Rise of Tomb Raider
Tomb Raider 2013
Witcher 3
Batman Arhkam Knight

Config:
Gigabyte P55 Ud3R
I7 870 OC 3.8Ghz
12 GB Consair DDR 3 1600 RAM
latest Nvidia game ready driver (For my Rise of Tomb Raider)
Zotac GTX 970 Stock speed

PS1: I have done IntelBurnTest before for 1 hour and windows did not crash, so I assume it has to do with my Graphics Card

PS2: I have had a clean installation of driver but still problem remains.

PS3: GPU temp is always kept at Max. 80C degree from Zotac Firestorm program.

Should I return the card when I still can?

Many Thanks!


 
Solution
I
Last Update:


@rcald2000, I have already done (1) and never done (3). for (2) sorry I didnt and I do not think it could make a difference because I have said I used an old GTX 460 and there was no more problem.

So dear all. I got a new GTX 970 from another brand (EVGA 970 FTW+). It requires x1 8pin and x1 6 pin power as compared to my previous 2X 6 pin Zoatac GTX 970.

Anyway, Good news! No more problem. I have played game for almost 7 hours, (3 hours consecutively) and there are no problems. No Glitches, no crash and when I check the Event Viewer, there are no more Nivida Display Driver has stopped and has been recovered shit.

Great! So i guess the conclusion is that my previous GTX 970 has a hardware problem (Game crash +...
Sort by date Sort by votes
Also, do you overclock your CPU? what is the blck value?
Sometimes a value different than 100 might put strain on the GPU and crash it.
I have this problem with a EVGA 550ti whenever i put different BLCK than 100.
 
Upvote 0 Downvote
May worth trying to run memtestCL. Its a OpenCL based graphics memory integrity checker for video cards. It's a less friendly than memtest86 that you're used to, but its helpful to identify video memory errors.

You can run it in Windows. Configure it to test as much of your memory as possible (try 3GB, if it says it can't allocate 3GB, try 2.5GB etc.) You can't allocate everything because running Windows itself will use some of it.

It's not ideal and nowhere near as friendly as memtest86, but it works in a pinch. I had faulty memory in my video card and it showed very similar symptoms to what you're describing.
 
Upvote 0 Downvote
1. how can I stress test a GPU for free? Any program?
2. the temp is max at 80 because the Zotac program will automatically control the fan. Of course I can increase the fan speed to lower the temperature.
3. Yes I overclock the CPU, BCLK is 160 (instead of 133), but I set PCI express speed to Auto.

Plus I have no problem with 3D Mark, no crash, no glitches....
 
Upvote 0 Downvote
all right. Will set the CPU to Stock (133 BCLK) and try again. Will report later.

@Amd4lif3. Thanks for the Intel tool, but you kept mentioning bottleneck, I dont see how bottleneck can crash a game, worse cast is I do not get enough performance what GTX 970 and truely deliver, but based on the 3D mark, I have compared my result with others, who have the same 970 but with i3770k, I am only about 1000 marks below.
 
Upvote 0 Downvote
@sancho_mic

I will set BCLK to 133 tonight. my i7 870 is old, but I have checked several benchmark, it doesnt affect much for gaming as I am playing at 2560X1080, it's more GPU bound then CPU bound.

Its 4 Core + HyperThread, stock speed is 3.2Ghz
 
Upvote 0 Downvote


i mention bottleneck not because of the power of the cpu but the drivers usually being strange with new gpus and old cpus they dont work right somecases so for example some times it may be freaking out in which case unpark the cpu
 
Upvote 0 Downvote
I hope your GPU instability will be solved by running clock speed. At least in my old pc this was the solution.

E.g. same pc overclocked was ok in w7 but once upgraded to win 10, i got stability issues.
I got annoyed when also in Linux (elementary os) i was facing instability, like system crashing with some GPU error.
So I decided to stop overclocking, returned fsb on 100 (or 133, not sure). This solved my problem.

Maybe it was not the optimal way to overclock that mobo and cpu, not sure.

If this solves you, i believe the GPU is quite inflexible and very picky on this kind of overclock.

You could try to increase gradually and see when/if it happens.

good luck!
-s
 
Upvote 0 Downvote
no luck. I have done the following.

1. GPU memory test using memtestG80 (3200MB, 100 Iterations) -- NO ERROR
2. NO MORE Overclocking of CPU, reset BCLK to 133, running at STOCK speed
3. Set the GPU fan to kick-in more aggressively, so now temperature is only about 6X max 70C degrees even at full load (I notice that when the game crashed, the temp of GPU is only around 68 too)

Still the game crashed. I checked the ERROR from Event viewer. (Event ID 1000)

Faulting application name: ROTTR.exe, version: 1.0.604.2, time stamp: 0x56a7981e
Faulting module name: ROTTR.exe, version: 1.0.604.2, time stamp: 0x56a7981e
Exception code: 0xc0000005
Fault offset: 0x0000000003220fec
Faulting process id: 0x1a20
Faulting application start time: 0x01d15aee4fe67e25
Faulting application path: Z:\SteamLibrary\steamapps\common\Rise of the Tomb Raider\ROTTR.exe
Faulting module path: Z:\SteamLibrary\steamapps\common\Rise of the Tomb Raider\ROTTR.exe
Report Id: fa962d6c-b2ea-418b-ab05-eca2d837cb99
Faulting package full name:
Faulting package-relative application ID:
PLUS, I found our that every time just before the game crash, there is a warning..

Display driver nvlddmkm stopped responding and has successfully recovered.
Event ID 4101

3. If I select "performance" in Geforce Experience, it seems able to last longer.. but result is not conclusive yet.


So I guess I need to solve this "Display driver nvlddmkm stopped responding and has successfully recovered." problem?

I am gonna do a MemTestX86+ just to make sure the memory is fine. ANy new idea?
 
Upvote 0 Downvote
hi all, another update.

1. I have tried updating the BIOS to the latest one for my motherboard, lol it wont even boot, so I revert back to old BIOS. no changed has been made to motherboard

2. Tried FurMark StressTest for 30 minutes, but do not seem to have problem.

3. I have ordered a replacement for my card. But meanwhile I want to show you guys the glitch I am having. I wanna know if Glitch = Hardware problem or could it be driver.

Conclusion: It will take few days for me to get a replacement, I plan to plug in my Old GTX 460 and see if I have this problem. If my old GTX 460 has no problem could it mean that it's a hardware failure for my new GTX 970?

This is an example of gltich I see once every 30 minutes. What do you think about this glitch?
lNBMLk5.jpg
 
Upvote 0 Downvote
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom