I've been having issues ever since upgrading my computer about a month ago with games crashing. I've upgraded my CPU, GPU, Mobo, and RAM intiially. Anytime I played any games (CS:GO, COD:WZ, Division 2) I would crash with Application has been blocked from accessing Graphics Hardware or the nvlddmkm has stopped responding and recovered. I did a DDU clean uninstall and install in Safe Mode in case the drivers were messed up, but still received that error. After this, I decided to clean install windows, in case DDU still hasn't fully wiped out the drivers somehow but still had the issues. I ran stress tests on everything I could think of to see if the crashes happen, but it seems to be limited to games at this point. I've ran Cinebench for CPU, MSI Afterburner/Furmark for GPU, and memtest for RAM, and all results were fine no crashes. At this point I was lost, and believed it may not have been software related. I googled and tried to do every fix I found. The fixes I can remember are TdrDelay at QWORD 8/TdrLevel 0, debug mode on NVIDIA Control Panel, Link State Power Management set to Off on Windows power options, Power Management Mode in NVIDIA Control Panel set to Prefer Maximum Performance, turn off NVIDIA in-game overlay, and re-seat the RAM/GPU. The change that seemed to help the most that I later found was actually underclocking the GPU. I initially do -20 MHz, which worked for about a week. Once I crashed once, I lowered it to -100 MHz, which worked for about another week. I then moved it to -200 MHz, and that worked for about a week again. That was when I decided it was time to try reaching out to EVGA support. I've explained everything, and eventually led to an RMA.
 
Fast forward to installing the new graphics card, I was able to play for about an hour before the issue occurred again. I decided maybe I should DDU and clean install again, so I did that in safe mode but it still did not help. However now, in addition to the previous 2 issues popping up in the Event Viewer logs, I am bombarded with a bunch of nvlddmkm Event ID 13 errors. I am currently sitting at -250 MHz on this new Graphics card and it is still happening. I began wondering if GPU is not the true issue, so I began testing my system with just a single RAM stick one at a time, and they both crashed individually.
 
If anyone has encountered this or has any information on how to fix this, I am open to trying anything at this point. I've exhausted almost all my options at this point, and am feeling like I just wasted money on my GPU.
 
\Device\Video3
Graphics SM Warp Exception on (GPC 3, TPC 0, SM 1): Out Of Range Register
 
\Device\Video3
Graphics Exception: ESR 0x51c7b0=0x3000d 0x51c7b4=0x0 0x51c7a8=0x7c12b72 0x51c7ac=0x174
 
\Device\00000094
Graphics Exception: ILLEGAL_OPCODE
 
\Device\00000094
Graphics Exception: ESR 0x404490=0x80000004
 
\Device\00000094
Graphics Exception: Shader Program Header 18 Error
 
System Spec:
CPU: Ryzen 5 3600X
RAM: G-Skill Trident Z 3200 CL16
Mobo: MSI B450 Tomahawk Max
GPU: EVGA 2070 Super XC Ultra Gaming
PSU: EVGA 850 GQ 80+ Gold
 
 
Thanks in advance!

post released from filter .. sorry for the delay

Time to pop the card in another known good working pc to find out what the real deal is.

Due to the whole pandemic and all, I have not had the chance to try that yet. I will try to get that done soon, but in the meantime are there anything else that I would be able to try? I can still mitigate the issue by underclocking a lot.
EDIT: It's also a bit weird how the first card I bought had the exact same issue as the RMA card.

I'm having the same problem, if you google around this is a big issue. I tried everything reinstalled and updated all drivers and windows and still getting crashes saying Application has been blocked from accessing Graphics hardware.  I had a AMD before and never had any of these issues.

This on a fresh re-format of the OS?

KingRemInTime
I'm having the same problem, if you google around this is a big issue. I tried everything reinstalled and updated all drivers and windows and still getting crashes saying Application has been blocked from accessing Graphics hardware.  I had a AMD before and never had any of these issues.

Are you seeing this for this card specifically? I've had all NVIDIA EVGA cards before this current 2070 Super and never had any issues. I've had to now underclock my gpu by -1000 MHz in order to even stop all crashing. The current underclock has prevented any crashes for a long while, but in the long run I don't think this is feasible as I paid for half the power essentially.


 

screwtech02
This on a fresh re-format of the OS?

Yes, it happened pre-reformat and post-reformat

Well, if your downclocking your card, and that seems to correct the issue, that leaves about 2 things, either the card, or the cooling on the card.  I'd take it back up with support and see what they say.

screwtech02
Well, if your downclocking your card, and that seems to correct the issue, that leaves about 2 things, either the card, or the cooling on the card.  I'd take it back up with support and see what they say.

I actually RMA'd my first card, and issue still occurred on the 2nd card. I haven't had a chance still to test on a different system since I gave my old parts to a friend, and don't want to go out too much during quarantine right now. In the meantime, if there are any other ways of confirming whether I just got extremely unlucky with 2 defective cards in a row, I'm all ears though.

tomwzhere

screwtech02
Well, if your downclocking your card, and that seems to correct the issue, that leaves about 2 things, either the card, or the cooling on the card.  I'd take it back up with support and see what they say.

I actually RMA'd my first card, and issue still occurred on the 2nd card. I haven't had a chance still to test on a different system since I gave my old parts to a friend, and don't want to go out too much during quarantine right now. In the meantime, if there are any other ways of confirming whether I just got extremely unlucky with 2 defective cards in a row, I'm all ears though.

What PSU are you using? Are you using 2 cables going to the card or 1 with 2 connectors? Downclocking the card that much is very strange. I see your original card did fail testing but not for what the RMA was started for.