Nvidia Geforce Drivers Release Announcement thread

Fortnite warned me on startup that I should not use Dx12 with the latest Nvidia driver, and lo and behold I crashed twice.

Ironically this is a way worse experience than I ever received on AMD, even at the height of their driver troubles.
 
Nvidia has another stab at fixing the driver issues that has been plaguing them, lots of lines in the fixed issues referring to black screen problems

Driver Version: 576.02 | WHQL
Release Date: Wed Apr 16, 2025
Operating System: Windows 10 64-bit,
Windows 11
Language: English (US)
File Size: 857.6 MB


Game Ready for GeForce RTX 5060 Ti

This new Game Ready Driver supports the new GeForce RTX 5060 Ti GPU and provides the best gaming experience for the latest new games supporting DLSS 4 technology including Black Myth: Wukong and No More Room in Hell 2.

Gaming Technology

  • Adds support for GeForce RTX 5060 Ti

Fixed Gaming and General Bugs

  • Please find an extensive list of fixed bugs in release note PDF (under Additional Information)
Cut and pasted from the PDF:

3.1 Fixed Issues in Version 576.02 WHQL
3.1.1 Fixed Gaming Bugs
> [Fortnite] random crashes during gameplay [5171520]
> [The First Berserker: Khazan] DXGI_ERROR_DEVICE_REMOVED Crash [5195216]
> [Star Wars Outlaws] Application will freeze after leaving the game idle for 5+ minutes
[5191099]
> Game stability issues when playing games with DLSS Frame Generation + GSYNC
[5144337]
> [Monster Hunter Wilds] Crash after accepting quest with DLSS-FG Enabled [5087795]
> [InZOI] Game crashes with error "GPU crashed or D3D Device Removed" [5154864]
> [Overwatch 2] Stutter when using VSYNC [5171856]
> [Hellblade 2 Senua's Saga] Increased aliasing when using TSR [5125662]
> [Hellblade 2 Senua's Saga] Crashing when using Smooth Motion [5209772]
> [The Last of Us Part 1] Crash when using Smooth Motion [5208799]
> Dithering/banding in some games on RTX 50-series GPUs [5121715]
> [Control] Flickering corruption in multiple areas [5118876]
> Stutter when using VSYNC [5202703][5202474]
> VSYNC in NVCP + frame generation causes issues in DLSS 4 games [5124816]
> [UNCHARTED: Legacy of Thieves Collection] Artifacts on screen when collecting
treasures [5158954]

3.1.2 Fixed General Bugs
> Stability issues when using Windows 11 24H2 [5160948]
Release 575 Driver for Windows, Version 576.02 RN-08399-576.02_v01 | 15
> Bugcheck w/ PAGE_FAULT_IN_NONPAGED_AREA (50) when playing games with DLSS 4 multi-frame generation [5144337]
> [RTX 50 series] GPUs crashes with black screen/underflow [5160845]
> [RTX 50 series] Random Black Screen issues [5090505]
> General system stability issues [4921925]
> [RTX 50 series] System hard hang with 572.16 driver loaded [5107271]
> Compute Shader related tests are failing due to "error" [4894179]
> [HWBU][DT GB202/203][LG 27GN950 Specific]: Display blacked out when applying 120Hz refresh rate [5044229]
> PC display will not wake after extended sleep time [5131052]
> Two DP output of the RTX5090 will blue screen when trying to watch protected videos [5167145]
> Black screen issue when testing 3D mark with driver 572.02 and 572.16 [5095825]
> Primary Blank display showing blank after hot plug the display in daisy chain [4978206]
> Display shows blank screen on setting RR 165/200Hz when daisy chain is enabled. [5049227]
> Second display showing blank when we apply higher RR for second display when displays connected in daisy chain [4956573]
> Primary monitor (AOROUS FO32U2P) goes blank when we HPD/power cycle second display in Daisy chain. [5075448]
> GeForce RTX 50 series GPUs crashes with black screen when playing graphically demanding games [5098914]
> RTX 50 series displays blank screens on LG 5k2k 45GX950A-B when running in DisplayPort 2.1 mode w/ HDR [5192671]
> Black screen on installing drivers and booting into Windows [5153642]
> DP2.1 - UHBR10/13.5 link rates are unstable on LGE 27GX790A-B [5080789]
> Multiple users reporting black screen issue when disable the "FCH Spread Spectrum" settings [5204493]
> [RTX 50 series] Slightly higher DPC latency may be observed on some system configuration [5168553]
> [RTX 50 series] Varjo Aero VR headset may fail to establish connection [5117518]
• Changing state of "Display GPU Activity Icon in Notification Area" does not take effect until PC is rebooted [4995658]
> [RTX 50 series] Display may show black screen when selecting DLDSR resolution [5144768]
> [Octanebench] Performance regression [3523803]
> [DaVinci Resolve] UI overlay in Fusion page is not displayed correctly [4974721]


 
Nvidia has another stab at fixing the driver issues that has been plaguing them, lots of lines in the fixed issues referring to black screen problems

Driver Version: 576.02 | WHQL
Release Date: Wed Apr 16, 2025
Operating System: Windows 10 64-bit,
Windows 11
Language: English (US)
File Size: 857.6 MB


Game Ready for GeForce RTX 5060 Ti

This new Game Ready Driver supports the new GeForce RTX 5060 Ti GPU and provides the best gaming experience for the latest new games supporting DLSS 4 technology including Black Myth: Wukong and No More Room in Hell 2.

Gaming Technology

  • Adds support for GeForce RTX 5060 Ti

Fixed Gaming and General Bugs

  • Please find an extensive list of fixed bugs in release note PDF (under Additional Information)
Cut and pasted from the PDF:

3.1 Fixed Issues in Version 576.02 WHQL
3.1.1 Fixed Gaming Bugs
> [Fortnite] random crashes during gameplay [5171520]
> [The First Berserker: Khazan] DXGI_ERROR_DEVICE_REMOVED Crash [5195216]
> [Star Wars Outlaws] Application will freeze after leaving the game idle for 5+ minutes
[5191099]
> Game stability issues when playing games with DLSS Frame Generation + GSYNC
[5144337]
> [Monster Hunter Wilds] Crash after accepting quest with DLSS-FG Enabled [5087795]
> [InZOI] Game crashes with error "GPU crashed or D3D Device Removed" [5154864]
> [Overwatch 2] Stutter when using VSYNC [5171856]
> [Hellblade 2 Senua's Saga] Increased aliasing when using TSR [5125662]
> [Hellblade 2 Senua's Saga] Crashing when using Smooth Motion [5209772]
> [The Last of Us Part 1] Crash when using Smooth Motion [5208799]
> Dithering/banding in some games on RTX 50-series GPUs [5121715]
> [Control] Flickering corruption in multiple areas [5118876]
> Stutter when using VSYNC [5202703][5202474]
> VSYNC in NVCP + frame generation causes issues in DLSS 4 games [5124816]
> [UNCHARTED: Legacy of Thieves Collection] Artifacts on screen when collecting
treasures [5158954]

3.1.2 Fixed General Bugs
> Stability issues when using Windows 11 24H2 [5160948]
Release 575 Driver for Windows, Version 576.02 RN-08399-576.02_v01 | 15
> Bugcheck w/ PAGE_FAULT_IN_NONPAGED_AREA (50) when playing games with DLSS 4 multi-frame generation [5144337]
> [RTX 50 series] GPUs crashes with black screen/underflow [5160845]
> [RTX 50 series] Random Black Screen issues [5090505]
> General system stability issues [4921925]
> [RTX 50 series] System hard hang with 572.16 driver loaded [5107271]
> Compute Shader related tests are failing due to "error" [4894179]
> [HWBU][DT GB202/203][LG 27GN950 Specific]: Display blacked out when applying 120Hz refresh rate [5044229]
> PC display will not wake after extended sleep time [5131052]
> Two DP output of the RTX5090 will blue screen when trying to watch protected videos [5167145]
> Black screen issue when testing 3D mark with driver 572.02 and 572.16 [5095825]
> Primary Blank display showing blank after hot plug the display in daisy chain [4978206]
> Display shows blank screen on setting RR 165/200Hz when daisy chain is enabled. [5049227]
> Second display showing blank when we apply higher RR for second display when displays connected in daisy chain [4956573]
> Primary monitor (AOROUS FO32U2P) goes blank when we HPD/power cycle second display in Daisy chain. [5075448]
> GeForce RTX 50 series GPUs crashes with black screen when playing graphically demanding games [5098914]
> RTX 50 series displays blank screens on LG 5k2k 45GX950A-B when running in DisplayPort 2.1 mode w/ HDR [5192671]
> Black screen on installing drivers and booting into Windows [5153642]
> DP2.1 - UHBR10/13.5 link rates are unstable on LGE 27GX790A-B [5080789]
> Multiple users reporting black screen issue when disable the "FCH Spread Spectrum" settings [5204493]
> [RTX 50 series] Slightly higher DPC latency may be observed on some system configuration [5168553]
> [RTX 50 series] Varjo Aero VR headset may fail to establish connection [5117518]
• Changing state of "Display GPU Activity Icon in Notification Area" does not take effect until PC is rebooted [4995658]
> [RTX 50 series] Display may show black screen when selecting DLDSR resolution [5144768]
> [Octanebench] Performance regression [3523803]
> [DaVinci Resolve] UI overlay in Fusion page is not displayed correctly [4974721]


It’s funny watching the same thing that happened 5-6 years ago with AMD play out with Nvidia.

1) driver has a clear amount of problems
2) people come out to claim there’s actually no problems, it’s all blown out of proportion by YouTubers/haters/whatever
3) IHV comes out and applies fixes for the exact problems described in 1), confirming it was a real problem and not just user error.

I am hoping this will fix my alt-tab crashes I’ve been having lately, it’s been driving me so nuts I almost just bought an AMD card just to escape it.
 
Nvidia has another stab at fixing the driver issues that has been plaguing them, lots of lines in the fixed issues referring to black screen problems

Driver Version: 576.02 | WHQL
Release Date: Wed Apr 16, 2025
Operating System: Windows 10 64-bit,
Windows 11
Language: English (US)
File Size: 857.6 MB


Game Ready for GeForce RTX 5060 Ti

This new Game Ready Driver supports the new GeForce RTX 5060 Ti GPU and provides the best gaming experience for the latest new games supporting DLSS 4 technology including Black Myth: Wukong and No More Room in Hell 2.

Gaming Technology

  • Adds support for GeForce RTX 5060 Ti

Fixed Gaming and General Bugs

  • Please find an extensive list of fixed bugs in release note PDF (under Additional Information)
Cut and pasted from the PDF:

3.1 Fixed Issues in Version 576.02 WHQL
3.1.1 Fixed Gaming Bugs
> [Fortnite] random crashes during gameplay [5171520]
> [The First Berserker: Khazan] DXGI_ERROR_DEVICE_REMOVED Crash [5195216]
> [Star Wars Outlaws] Application will freeze after leaving the game idle for 5+ minutes
[5191099]
> Game stability issues when playing games with DLSS Frame Generation + GSYNC
[5144337]
> [Monster Hunter Wilds] Crash after accepting quest with DLSS-FG Enabled [5087795]
> [InZOI] Game crashes with error "GPU crashed or D3D Device Removed" [5154864]
> [Overwatch 2] Stutter when using VSYNC [5171856]
> [Hellblade 2 Senua's Saga] Increased aliasing when using TSR [5125662]
> [Hellblade 2 Senua's Saga] Crashing when using Smooth Motion [5209772]
> [The Last of Us Part 1] Crash when using Smooth Motion [5208799]
> Dithering/banding in some games on RTX 50-series GPUs [5121715]
> [Control] Flickering corruption in multiple areas [5118876]
> Stutter when using VSYNC [5202703][5202474]
> VSYNC in NVCP + frame generation causes issues in DLSS 4 games [5124816]
> [UNCHARTED: Legacy of Thieves Collection] Artifacts on screen when collecting
treasures [5158954]

3.1.2 Fixed General Bugs
> Stability issues when using Windows 11 24H2 [5160948]
Release 575 Driver for Windows, Version 576.02 RN-08399-576.02_v01 | 15
> Bugcheck w/ PAGE_FAULT_IN_NONPAGED_AREA (50) when playing games with DLSS 4 multi-frame generation [5144337]
> [RTX 50 series] GPUs crashes with black screen/underflow [5160845]
> [RTX 50 series] Random Black Screen issues [5090505]
> General system stability issues [4921925]
> [RTX 50 series] System hard hang with 572.16 driver loaded [5107271]
> Compute Shader related tests are failing due to "error" [4894179]
> [HWBU][DT GB202/203][LG 27GN950 Specific]: Display blacked out when applying 120Hz refresh rate [5044229]
> PC display will not wake after extended sleep time [5131052]
> Two DP output of the RTX5090 will blue screen when trying to watch protected videos [5167145]
> Black screen issue when testing 3D mark with driver 572.02 and 572.16 [5095825]
> Primary Blank display showing blank after hot plug the display in daisy chain [4978206]
> Display shows blank screen on setting RR 165/200Hz when daisy chain is enabled. [5049227]
> Second display showing blank when we apply higher RR for second display when displays connected in daisy chain [4956573]
> Primary monitor (AOROUS FO32U2P) goes blank when we HPD/power cycle second display in Daisy chain. [5075448]
> GeForce RTX 50 series GPUs crashes with black screen when playing graphically demanding games [5098914]
> RTX 50 series displays blank screens on LG 5k2k 45GX950A-B when running in DisplayPort 2.1 mode w/ HDR [5192671]
> Black screen on installing drivers and booting into Windows [5153642]
> DP2.1 - UHBR10/13.5 link rates are unstable on LGE 27GX790A-B [5080789]
> Multiple users reporting black screen issue when disable the "FCH Spread Spectrum" settings [5204493]
> [RTX 50 series] Slightly higher DPC latency may be observed on some system configuration [5168553]
> [RTX 50 series] Varjo Aero VR headset may fail to establish connection [5117518]
• Changing state of "Display GPU Activity Icon in Notification Area" does not take effect until PC is rebooted [4995658]
> [RTX 50 series] Display may show black screen when selecting DLDSR resolution [5144768]
> [Octanebench] Performance regression [3523803]
> [DaVinci Resolve] UI overlay in Fusion page is not displayed correctly [4974721]


Ah! A bug fix list almost a page long, feels like AMD. :)
 
I found this one interesting:

"Multiple users reporting black screen issue when disable the "FCH Spread Spectrum" settings [5204493]"

That's normally a motherboard setting to reduce radiated EMI and interference, or at least reduce the peak amplitude of, by modulating the clock speed of the component in question about its intended target frequency.

The component they're talking about in this case though, the FCH, I think is AMD specific terminology, for 'Fusion Controller Hub' which is now part of the CPU die (or rather, the I/O chiplet for desktop CPUs) I wonder if the black screens, or at least some of them, are related to PCIe signal integrity being on the ragged edge and picking up some radiated EMI from somewhere? The mass proliferation over the last half decade or so of PCIe riser cables for horizontal GPU placements probably aren't helping matters on that front, especially with the transition to PCIe 4.0 and 5.0.

Would be interesting to see if some folks affected by the black screen issue ended up with some relief falling back to PCIe 4.0 or even 3.0, if only just for testing purposes.
 
3) IHV comes out and applies fixes for the exact problems described in 1), confirming it was a real problem and not just user error.
No one ever said all of it was "user error". The amount of issues the previous driver had was overblown though, and this bugfix list actually proves it - if you read through it carefully you'll notice that about 2/3rds of bugs listed there are actually just a couple of bugs which were fixed with the new driver. Namely - bugchecks with Gsync+FG, stuttering issues with vsync+Gsync and another (third?) round of "black screens" fixing which are most likely related to the new DP 2.1 implementation. The first two were driver issues but these were only two issues really. Everything else was (and likely still is) system specific and not necessarily related to the driver.

I've had just one issue in one game with the previous driver which this driver fixed. Hardly a "load of" fixes.
 
Nvidia should base their debugging efforts based solely on Degustor's personal experience. No need for all those fixes because he only had one issue in one game.
 
Nvidia should base their debugging efforts based solely on Degustor's personal experience. No need for all those fixes because he only had one issue in one game.
Nvidia's debugging efforts hasn't changed and the amount of issues the drivers have is the same as it has been over the last years.
The gaslighting about the recent drivers being somehow especially bad in this needs to stop.
 
Nvidia's debugging efforts hasn't changed and the amount of issues the drivers have is the same as it has been over the last years.
The gaslighting about the recent drivers being somehow especially bad in this needs to stop.
You can continue to deny it all you want but it doesn't change the real world, there's no gaslighting about the drivers.
NVIDIA just posted drivers with literally the longest list of fixes ever in a single NV driver, how could that happen when the amount of issues is the same as it has always been over the last years?
And no, based on user reports the new drivers didn't fix all the issues, not even all of those the drivers claim to have fixed.
 
Nvidia's debugging efforts hasn't changed and the amount of issues the drivers have is the same as it has been over the last years.
The gaslighting about the recent drivers being somehow especially bad in this needs to stop.
No one ever said all of it was "user error". The amount of issues the previous driver had was overblown though, and this bugfix list actually proves it - if you read through it carefully you'll notice that about 2/3rds of bugs listed there are actually just a couple of bugs which were fixed with the new driver. Namely - bugchecks with Gsync+FG, stuttering issues with vsync+Gsync and another (third?) round of "black screens" fixing which are most likely related to the new DP 2.1 implementation. The first two were driver issues but these were only two issues really. Everything else was (and likely still is) system specific and not necessarily related to the driver.

I've had just one issue in one game with the previous driver which this driver fixed. Hardly a "load of" fixes.
Brother I haven’t seen a bug fix list this big on an Nvidia driver release literally ever.

Like I said, exact same thing played out on the AMD side years ago: they had a clear problem, the usual suspects said there’s actually no problem, IHV fixed the problem and the usual suspects switched from “it’s not real” to “okay it was real but now it’s fixed!”
 
Brother I haven’t seen a bug fix list this big on an Nvidia driver release literally ever.

Like I said, exact same thing played out on the AMD side years ago: they had a clear problem, the usual suspects said there’s actually no problem, IHV fixed the problem and the usual suspects switched from “it’s not real” to “okay it was real but now it’s fixed!”

It's ok to discuss about possible quality decline of NVIDIA's driver but there's no need to exaggerate. NVIDIA did (probably still do) publish PDF release notes and they used to have multiple pages of bug fixes. It's not even that long ago.
 
You can continue to deny it all you want but it doesn't change the real world, there's no gaslighting about the drivers.
You're quite literally gaslighting right now because in the "real world" all the awful numerous issues has seemingly been fixed by just one driver update. Which means that there weren't a lot of them to begin with.

NVIDIA just posted drivers with literally the longest list of fixes ever in a single NV driver, how could that happen when the amount of issues is the same as it has always been over the last years?
A new GPU family launch always tend to produce an increase in the amount of driver issues which are then being fixed. What I can't understand is how people suddenly seem to be unaware of that now in 2025.
And the list is both misleading (it lists issues and not actual fixes done to these issues of which there are way less than issues) and likely inflated purposefully so that those who were running around the internet screaming can now say "see, they've fixed all of these issues".

And no, based on user reports the new drivers didn't fix all the issues, not even all of those the drivers claim to have fixed.
Which again points us to the fact that many issues people are having aren't in fact driver induced or even related.

Brother I haven’t seen a bug fix list this big on an Nvidia driver release literally ever.
See above.
 
It's ok to discuss about possible quality decline of NVIDIA's driver but there's no need to exaggerate. NVIDIA did (probably still do) publish PDF release notes and they used to have multiple pages of bug fixes. It's not even that long ago.
I can’t remember the last time that happened.

You're quite literally gaslighting right now because in the "real world" all the awful numerous issues has seemingly been fixed by just one driver update. Which means that there weren't a lot of them to begin with.


A new GPU family launch always tend to produce an increase in the amount of driver issues which are then being fixed. What I can't understand is how people suddenly seem to be unaware of that now in 2025.
And the list is both misleading (it lists issues and not actual fixes done to these issues of which there are way less than issues) and likely inflated purposefully so that those who were running around the internet screaming can now say "see, they've fixed all of these issues".


Which again points us to the fact that many issues people are having aren't in fact driver induced or even related.


See above.
To preface: gaslighting is a term describing a thing abusive spouses use to coerce their wives. The fact that it’s entered the general lexicon and we’re using it to defend a multi billion dollar company’s drivers feels a bit overdramatic to me.

I’m not really sure what you’re arguing. People have been saying for weeks that this round of drivers had a lot of bugs. Game developers have co-signed this (Fortnite showed a warning to me saying to either use DX11/performance mode or (implicitly) downgrade my drivers, unless you think Epic Games have no clue what they’re talking about). Nvidia released a driver that says they fixed many of the issues we’ve been having. Which part of that do you disagree with?
 
Part of it could be that some people have this innate desire to defend corporations (and it appears to be a mostly American phenomena) I'm quite surprised no one has posted "It's your own fault for buying Nvidia"
 
This dialog got me curious, so I scrolled back through this thread over the past -- I dunno, 15 or 20 minutes -- clicking on the linked driver release PDF notes. Were there older drivers with longer lists of bugs (either fixed, or unfixed but known?)

The longest list I found in my cursory glance over maybe a dozen pages of this thread was back in October '22; interestingly enough linked to the initial ADA release. I suspect new graphics chip generations bring out the worst in a driver, which shouldn't be a surprise to anyone who has ever written code in their lives. Anyway, here's the linked PDF from that driver: https://us.download.nvidia.com/Windows/526.47/526.47-win11-win10-release-notes.pdf Still seems like a "less long" list than we're looking at today.

At the same time, for those who have never written code in your lives, it may surprise you to know that one singular bug can manifest in a lot of seemingly different ways. Think about it in terms of sickness symptoms -- you can have ONE viral infection which manifests in a dozen different physical ailments. Doesn't mean a dozen things are wrong with you,,,
 
I think people tend to exaggerate when the bugs they encounter is significant to their use case or don't have obvious workaround. For example, I also encounter multiple bugs using an NVIDIA GPU right now, such as infrequent HDR flash during some video playback. But some are not even introduced recently. To me I don't feel a significant worsen of driver quality. Each new driver version might introduced some new bugs while fixing other. For others maybe, but saying this is "the worst ever" seems to me an exaggeration unless you never lived through the "Detonator" days.
 
At the same time, for those who have never written code in your lives, it may surprise you to know that one singular bug can manifest in a lot of seemingly different ways. Think about it in terms of sickness symptoms -- you can have ONE viral infection which manifests in a dozen different physical ailments. Doesn't mean a dozen things are wrong with you,,,
I feel this is relatively inconsequential to the end consumer: if they pay $1000 for a GPU they want VRR to work and they want Fortnite to not crash, regardless if it's one bug causing 99 issues or 99 bugs causing 1 issue.
 
Back
Top