Profile
PNG, GIF, JPG, or BMP. File must be at least 160x160px
and less than 600x600px.
Top Communities
- GeForce Graphics Cards
- GeForce Laptops
- G-SYNC Monitors
- Gaming PCs
- SHIELD TV
- SHIELD Tablet
- Virtual Reality
- Legacy Products
- Game-Ready Drivers
- GeForce Experience
- Instant Replay Recording
- Photo Mode
- Freestyle Filters
- Broadcasting
- GameStream
- GFN - General Chat
- GFN - Announcements
- Forum Feedback
- Off-Topic
- Creative Apps
- Games
- RTX Technology (DLSS & DXR)
- Highlights
- Studio Drivers
- 3D Vision
- GFN - Game Requests
- GFN - Looking for Group
- GFN - Tech Support
- NVIDIA Reflex
- NVIDIA App
rhylos
User
113
Total Comments
3
Total Topics
Any hints as to the known ones? Ref: "When using certain 3rd party performance overlays alongside DLSS Frame Generation, crashes can occur."
TL;DR Win11 24H2 KB5052093 may have got me operational again with 560.81 drivers. I believe I finally got my system back up and running without random black screens (GPU disconnect/power offs) after using multiple versions of the 572 series from .60 and below. Even falling all the way back to 560.81 drivers without GeForce Experience or the Nvidia app (just drivers). Many attempts for a system restore failed me. Even rolled back AMD Chipset drivers but I continued to have the black screen errors after installing the 572's. Was about to restore from backups software but it was older than I wanted. Microsoft then pushed out KB5052093 on me. After a reboot, I was finally, knock on wood, after to play VR games without the GPU crashing as well as no random black screens. Will provide feedback if I do get black screens in the next day or so after a few more reboots. Then I'll try 566.36 drivers again. Win11 24H2 AMD 7950x w/iGPU enabled and HDMI connected to a monitor ASUS TUF 4090 HDMI connected to LG C1 and DP connected to PSVR2 adapter and HMD ASRock Taichi x670e 32GB RAM OS SSD
Same
This is when we find out the driver update harmed our GPUs. Rolling back isn't helping anymore. Any real load placed on my GPU now turns it off while in VR. Kills everything connected to the card. C1, no connection till I power off the system completely or it remembers that the card is offline.
Keep an eye on your system, after 572.60, rolling back to 566.36 I was still having issues with black screen out of the blue via HDMI on the 4090. I'm all the way back to 560.81 now with Windows Drivers prevented through Windows Update for the interim.
4090 GPU Crash with these drivers. --In general STAY AWAY-- After a few reboots of seemingly normal operation, these drivers caused a hard crash of the 4090 GPU. Seemingly as though the card was removed from the system while in the OS. If it were not for AMDs CPUs integrated GPU (iGPU) I would have thought the system frozen. The system fell back to sending video output to the secondary display directly connected to the motherboard/CPU out HDMI port. There appear to be minimum to no useful information from the event log. Nor a bluescreen crash dump. I did manually perform a kernel dump while the system was still on, I fear it is a mini-dump though and may not have a useful cause. No games were running nor loaded. The Steam interface was running. Windows Key, CTRL, SHIFT +B did not recover the NVIDIA GPU to a working state. Windows display settings doesn't even show the GPU nor display options. The resolution and refresh rate section are grayed out. No action can be done there in this state. I disabled, the NIC, configured the system to boot into safe mode, DDU'd, and rebooted into normal mode and installed 566.36 for now. I believe Windows tries to install this build, I cannot remember, if I install an older driver version shortly afterwards. Windows 11 24H2 build 26100 ASRock x670e Taichi AMD 7950X w/iGPU enabled (Not OC'd) 32GB RAM OS on SSD ASUS TUF 4090 No custom global option selected except, prefer max power, g-sync windowed and full, and limit fps to 116) Windows display settings set to use 3840 res (4k), VRR, and 10bit, 120hz. (60hz causes tearing in COD even with g-sync on without vsync) HAGS On Game mode On HDR disabled AMD Chipset Driver: 7.02.13.148 Main display (LG C1) connected to 4090 via HDMI PSVR2 via DP via PSVR2 adapter. VR HMD powered off during session and incident. Secondary display (generic 1080 LCD, occasionally used with AIDA64 monitoring - monitoring off during session and incident) via HDMI port.
After a clean driver install, same issue with the force shutdown via power button. Even worse, I could not even do a ctrl alt del to see what was going on resource wise after reboot Forced reboot multiple times. Same issue. Could not get back into system. Did the three shutdown or restart option to get into Recovery Options and to Safe mode to perform a DDU, disable the active NIC and Wi-Fi so I could reboot into normal mode. Got in just fine, then installed 566.03. Issues appear gone. OS: Win 11 24H2 26100.3194 w/ WFEP 1000.26100.48.0 AMD 7950X plugged into secondary extended monitor via HDMI 32GB RAM SSD OS HD ASUS TUF 4090 plugged into LG C1 via HDMI and PSVR2 via DP
Did the driver get pulled from the Manual Downloads results for 4000 series and Win 11? Only seeing 572.16 and below listed for the 4090 (even with cache clears). Edit: 572.42 is showing up now. Interesting that the linked page does not state to use the clean install option to support [5081921] as this page does.
This fixed the error preventing the packaged version of ChatWithRTX from running; however, is there any fix or update that can correct the remaining notice on launch? "You try to use a model that was created with version 2.5.1, however, your version is 2.2.2. This might cause unexpected behavior or errors. In that case, try to update to the latest version."
@oguzhankurt, as if the VRAM is overclocked and showing crazy colors briefly on the screen? Are there tools to detect such an issue in order to send something to Nvidia to debug?
Integrate functions such as the Nvidia Inspector with reset to defaults options.
Brightness or Gamma auto correction feature for when VRR G-sync kicks in. This is to prevent the appearance of brightness fluctuations from monitors or displays (especially on OLED based displays).
Releases of HF drivers are appreciated when there is a good set of QA tests performed. Since the HF's are intended for those affected by a specific resolved issue, these particular updates are optional for use. I would welcome more of these between full release "rollups".
Did ver 537.09 (hotfix) have this problem as well?
Latency looks really good as well on my end. So far, this one is a keeper.