Question RTX 2070 causes system crashes when installing any NVIDIA drivers

Apr 30, 2019
6
0
10
0
I have been tearing my hair out for last few hours trying to figure this out, was hoping y'all could help a brother out.

I just got a new rig a couple weeks ago, built it myself (Specs below for troubleshooting), and now its acting up. About 5 hours ago, my screen suddenly when white, froze everything up, even the indicators for my keyboards Num Lock, Caps Lock, and Scroll Lock keys would not respond when I pressed them. Then the display recovered...until it straight up crashed the whole computer, no BSOD.

The computer restarted, but the screen was filled with artifacts, and the screen was so pixelated I had to use windows Magnifier tool to see any text on the screen. I clicked around and soon found out that the Nvidia control panel didn't appear when I right-clicked my desktop, but GeForce Experience still existed on my desktop. I managed to reinstall the latest drivers. Everything seemed ok until shortly after I logged in when my screen turned a solid shade of yellow, froze, then recovered. It alternates back and forth between my fully responsive desktop, and solid shades of color until it fails to recover or I turn to computer off. The colors so far have been: Yellow, White, Black, and Blue.

To diagnose the problem, I started in Safe Mode With Networking, which resolved the issues, but I soon found out that my card now had no drivers, but starting back in normal mode still resulted in the same issues! So I installed DDU and cleared all the BS out. I then tried installing a previous driver version (from March 25th, 2019, as I had the computer during the time which those drivers were installed, and had no issues). Every time I attempted to install the NVIDIA drivers, the problems reappeared. Eventually, I just DDU'd my system, and didn't reinstall any drivers. Now everything is fine, except my ~$500 GPU is worthless because my PC is using Windows Basic Display Adapter, not my GPU.

SPECS:
CPU: i7-8700 (NOT UNLOCKED)
GPU: Asus - GeForce RTX 2070 8 GB
STORAGE: Samsung 970 EVO Pro (1TB M.2 NVMe)
PSU: CORSAIR RMx Series RM1000X 1000W 80 PLUS GOLD
CPU COOLER: Coolermaster 212X
RAM: CORSAIR Vengeance LPX 32GB (2 x 16GB) (DDR4 3000)

I could really use some help on this from y'all. If my card is faulty I can RMA for a new one, but I'd rather try to fix what I have and go through other options before I do that.

Thanks in advance for all your help!
 

Tolis_GR

Reputable
Mar 13, 2017
1,527
20
4,165
317
I suspect your gpu has unstable vram. Try underclocking core and memory clocks just to check if the gpu has poor bios settings or hardware issues. If you can try your gpu in another computer that would help isolate the problem.
Some other things to check: Check your psu connections. Try another 6+2 pin connector. Load motherboard bios defaults. Try another pcie port
 
Apr 30, 2019
6
0
10
0
I suspect your gpu has unstable vram. Try underclocking core and memory clocks just to check if the gpu has poor bios settings or hardware issues. If you can try your gpu in another computer that would help isolate the problem.
Some other things to check: Check your psu connections. Try another 6+2 pin connector. Load motherboard bios defaults. Try another pcie port
Thank you for all your suggestions. I can only change the clock settings of the gpu if I have drivers, which causes the display to bug out, and I have MSI afterburner, so its possible if I do it when I have display. However, once I do get to MSI, it shows my memclock is 7000! I don't know too much about this stuff, but I do know that 7000 is not realistic, even at the highest end of systems. So you were correct, my memory is unstable. However, I can only drop it by 500, and it still remains unstable, only dropping to 6500ish. Can you tell me how to stabilize my memory? Is my card busted? How do I return my card to normal? Can I fix it without reseating my card and changing connectors?

Thank you for all your help so far Tolis_GR!
 

Tolis_GR

Reputable
Mar 13, 2017
1,527
20
4,165
317
MSI obviously reports the effective vram frequency. For GDDR5 7000 mhz effective is 1750 mhz true frequency, nothing wrong here. a fresh install always helps to delineate some issues but to me it looks you were unlucky with the gpu. Enough RTX cards had issues during launch
 
Apr 30, 2019
6
0
10
0
MSI obviously reports the effective vram frequency. For GDDR5 7000 mhz effective is 1750 mhz true frequency, nothing wrong here. a fresh install always helps to delineate some issues but to me it looks you were unlucky with the gpu. Enough RTX cards had issues during launch
So do I have to rma my gpu? Thats fine if I have to, but its not optimal. Is there any way to fix this on a software level?

Thanks for the info about how msi afterburner reports frequency, that's helpful to know.
 
Apr 30, 2019
6
0
10
0
MSI obviously reports the effective vram frequency. For GDDR5 7000 mhz effective is 1750 mhz true frequency, nothing wrong here. a fresh install always helps to delineate some issues but to me it looks you were unlucky with the gpu. Enough RTX cards had issues during launch
Also, it should be worth noting that I get artifacting when I boot. Also, reinstalling windows is giving me trouble because it installs gpt but windows needs mbr.

EDIT: It only artifacts sometimes on boot, about half the time, but it continues to the OS too.
 

TJ Hooker

Illustrious
Herald
Also, it should be worth noting that I get artifacting when I boot. Also, reinstalling windows is giving me trouble because it installs gpt but windows needs mbr.

EDIT: It only artifacts sometimes on boot, about half the time, but it continues to the OS too.
It artifacts in BIOS/prior to Windows loading? Or only after Windows starts loading?
 

ASK THE COMMUNITY

TRENDING THREADS