Hi all,
The Unreal Engine crash - NVBug #4769138 is also fixed. The fix will be available on the next 560 driver release.
Other issues remain under investigation.
Thanks!
Hi all,
The Unreal Engine crash - NVBug #4769138 is also fixed. The fix will be available on the next 560 driver release.
Other issues remain under investigation.
Thanks!
@jonathanraypollard, Thank you for the feedback. would you mind posting the crash stack trace from your systems for 560.28.03 driver?
On a Razer Blade 2021 Advanced (RTX3070 Max-Q) with Optimus hybrid graphics (Intel), the GPU never seems to go to sleep, even when in power save mode and no external display connected, just browsing the web.
This makes the battery melt like ice. Doesnât happen with Nouveau drivers.
Running Ubuntu 24.04, Open Drivers. Nvidia-smi reports at minimum 11W consumption.
Are there any news on HDR? There is a lot of feedback about washed out colors with a lot of monitors. AMD GPUs work excellent with HDR.
Hi @lord, Sorry I have not filed a bug for the HDR issue yet.
With regards to HDR. I have not yet had a chance to test 560 driver. But maybe this is helpful to document as an additional monitor.
555 with Samsung QN90B I cannot get any HDR from the cards display port. That is using a DP->HDM2.1 adapater. I tried 2 adapters and neither worked. I have a suspicion it is not an adapter issue though.
On HDMI 2.1 direct I can get it to work since 555, but Colour Space will only go to BT.709, which I believe is not correct and I do think certain games are washed out.
Hi! Please fix xwayland windows. Vram leaks when resizing windows. Exit session fixâŚ
Tested on rtx2060 and 4070. Fedora 40, kde and gnome. Nvtop in terminal for monitoring.
Driver 555 and 560.
Thanks for the reply! As far as I understand, the problem with Nvidia and Intel GPU is widespread: HDR and color management in KWin, part 3 | Xaverâs blog
the
Colorspaceproperty implementations from Intel and NVidia cause washed out colors on many displays, because the API requires the compositor to change the property value depending on whether or not communication with the display uses RGB or YUV encoding⌠which the compositor doesnât actually know anything about. The AMD implementation works around this by translating the property to the correct value in the kernel
In my case with both 555 and 560 monitor is not switching from BT.709. It correctly detects enabled HDR, but colors are washed out. There is a topic from man with Alienware AW3423DWF: Reddit - Dive into anything
There is also a bug ticket in KDE bug tracker with the exact problem: 482780 â Incorrect color gamut with HDR enabled
And members from KDE are also saying itâs a problem at Nvidia side.
I wrote my configuration previously here: 555 release feedback & discussion - #140 by lord
The only change is that now I use the latest Linux 6.10.2.
Ok, 560.31.02 is available for download at Linux x64 (AMD64/EM64T) Display Driver | 560.31.02 | Linux 64-bit | NVIDIA
For now, games are working ok for me (gnome+wayland+4070)
No fix for excessive XWayland memory usage? Or the multiple XID 109 bugs?
Can you check if Minecraft 1.12.2 dosnt crash on full screen. Thanks.
Yes, its fixed.
Thank you!
Just want to report with 560.31.02 and open kernel modules the issue is still there with direct scanout on multiple screens, i dont think that was fixed in this release, but still here is the log.
nvidia-bug-report.log.gz (774.7 KB)
With the 555 and 560 drivers, Iâm getting a consistent wine-vulkan assertation error (â!status && âvkAcquireNextImageKHRââ). Apparently, this occurs only in Wayland regardless of the distro used, and occurs only with the linux nvidia driver. Iâve seen reports that reverting back to version 545 seems to solve the problem, thatâs not possible for me as I need explicit sync and VRR for my older GPU.

Still no multi-monitor VRR folks.
thanks for the input. Looking like itâll be in 565 / 570 then ig.
I guess it will be mentioned when it is fixed. @abchauhan does this have a filed bug number?
Hi @Sanx1 , Yes, that is correct. The direct scanout issue is fixed but it did not make the cut-off for 560.31.02. We are verifying the fix on our systems. It should make it to the next release if all tests pass. Thank you.
Surely⌠Next version⌠Is there a bug filed for this I can follow?