• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

The AMD Driver Thread

I see. I don't suppose you have a picture of the actual error dialog? Never heard of anything like this happening so far

Sorry, I didn't grab a screenshot, I was just thinking to fix it. I'll try and remember to get it next time and revisit the issue.

I only got the error with this latest driver. With the previous driver there was no error message, just an extended blank screen during driver installation, before it came back. Once the screen never came back at all and a reset of the machine was required.
 
Sorry, I didn't grab a screenshot, I was just thinking to fix it. I'll try and remember to get it next time and revisit the issue.

I only got the error with this latest driver. With the previous driver there was no error message, just an extended blank screen during driver installation, before it came back. Once the screen never came back at all and a reset of the machine was required.
I see. The extended display PNPs (somewhere like 5mins long) really do need to be investigated. that's terrible UX. Could be related to what you saw in some way.
If you do hit that error dialog again, please try to grab a pic with your phone or something, would love to take a look
 
hey @AMD_Vik Any chance the Second Life flickering got fixed in this latest driver? I'm holding off until I hear some good news, cheers :)
I think a fix has been identified in the shader compiler but i need to work to getting that prioritised and aligned to a release. It's presently being pushed out due to its risk level
 
For what it's worth, I use Fancontrol to control my 9070XT fan. The most that happens when I install a new driver via the AMD update manager without restarting is that the Fancontrol GPU temp sensor goes to zero and the fans are no longer controlled by it. Which in my opinion is acceptable.

A reboot feels like a normal thing to do manually even if a driver update doesn't ask for it. That fixes it for me.
 
That doesn't sound like the issue we addressed. Can you try launching the game with igfx disabled and see what happens?
So I haven't been able to test myself yet due to work etc, however looking on Reddit, it seems as though it could be when using igpu, something I can't live without sadly.
 
Last edited:
So I haven't been able to test myself yet due to work etc, however looking on Reddit, it seems as though it could be when using igpu, something I can't live without sadly.
that's understandable. it's a valuable datapoint. please let me know if you get the chance to confirm
 
that's understandable. it's a valuable datapoint. please let me know if you get the chance to confirm
I can confirm that disabling iGPU allows the game to run fine, as soon as I reenable it the problem returns. If you could feed this back to your team I would be most grateful as there must be a fair few of us
 
I can confirm that disabling iGPU allows the game to run fine, as soon as I reenable it the problem returns. If you could feed this back to your team I would be most grateful as there must be a fair few of us
Can you tell me which CPU this is with?
 
good to know. looks like the internal ticket was rejected as 'not fixed' - the engineers there have already come to the conclusion about the iGPU causing persistent issues even with the AMF component fix, so thankfully that investigation is already underway. Will do what I can to make sure this is prioritised
 
Why am I still seeing things about how nvidias frequent driver updates make them better than AMD? My adrenaline has updates maybe twice per month. Isn't that enough? And I have never had to roll thme back because of instability or anything like I did for well over a year with my gtx970...
 
Last edited:
If that is true then people need a word with themselves. Reason nowadays for Nvidia releasing so many drivers is they just can't get a stable driver out the door. Got to love that advice for team green, "safer to choose a driver version that’s one or two releases behind the latest". Ouch.

 
Why am I still seeing things about how nvidias frequent driver updates make them better than AMD? My adrenaline has updates maybe twice per month. Isn't that enough? And I have never had to roll thme back because of instability or anything like I did for well over a year with my gtx970...
Radeon drivers used to have a bad rep years ago but they have been pretty good for years now still get the odd problem from time to time but over all they are really stable, But you still get people spouting the AMD drivers suck NVIDIA are so much better but that's mostly from people that have never owned an AMD GPU lol
 
Radeon drivers used to have a bad rep years ago but they have been pretty good for years now still get the odd problem from time to time but over all they are really stable, But you still get people spouting the AMD drivers suck NVIDIA are so much better but that's mostly from people that have never owned an AMD GPU lol
I've had a BSOD just from updating Adrenalin drivers last year which I've never had in many years of owning Nvidia cards. A few days ago I kept getting a black screen after re-applying a conservative undervolt in Adrenalin after updating the drivers and I had to do a force shutdown, the only way I found of re-applying my undervolt was to change the settings and apply them one at a time. I was unable to play FH6 for the first couple of weeks due to major stuttering issues on AMD GPUs, part of the reason I got the £110 Premium version was the 5 day early access. The last one wasn't technically an AMD driver fault, but it shows that you're going to get more issues with AMD because developers will prioritise testing the most popular GPUs which are Nvidia. I'll definitely be moving back to Nvidia when the next gen cards are released due to drivers.
 
I've had a BSOD just from updating Adrenalin drivers last year which I've never had in many years of owning Nvidia cards. A few days ago I kept getting a black screen after re-applying a conservative undervolt in Adrenalin after updating the drivers and I had to do a force shutdown, the only way I found of re-applying my undervolt was to change the settings and apply them one at a time. I was unable to play FH6 for the first couple of weeks due to major stuttering issues on AMD GPUs, part of the reason I got the £110 Premium version was the 5 day early access. The last one wasn't technically an AMD driver fault, but it shows that you're going to get more issues with AMD because developers will prioritise testing the most popular GPUs which are Nvidia. I'll definitely be moving back to Nvidia when the next gen cards are released due to drivers.
I think you may have other issues. Personally, I don't recall ever seeing a BSOD updating AMD drivers. That's thinking back to ye olden Vega64 days, to 6700XT, to my current 6900XT. On Windows 10 and now Windows 11. If the undervolt is giving you a black screen then that would point to the undervolt not being stable. That's not an AMD driver issue - that could just be down to poor silicone. You are not going to get "more" issues with AMD vs Nvidia. All I can safely say is, go have a read of the Nvidia drivers thread, do some reading on XDA, and those sorts of sites, and make up your own mind whether you think Nvidia drivers seem stable. :)
 
I wanna cap this off by saying, BSODs in the field are very real, for all vendors. Software will never not have bugs, and graphics drivers are one of the most difficult domains in all software engineering. If there are any reproducible cases for these, I need to know about em, and we'll do what we can to get them addressed.

Also want to clarify the remark about custom tuning - this is not a guarantee between drivers. If a given profile works at one point and no longer works at another, that's not necessarily a marker for overall stability. It's not uncommon for V/F to be tweaked with power management firmware updates over time to see better gfx utilisation for released GPUs. Vega was a key example of this; where it launched with a lot of tuning headroom, but eventually saw greater utilisation over time with sw updates (at the cost of said tuning headroom). We recently transitioned to a new IP branch, as such, all custom board tuning should be validated.
 
Not to derail further, but just to backup my point with some evidence (this is by no means complete).

1. GeForce 595.59 WHQL — Pulled After Widespread Breakage

Issue type: Severe stability problems
Reported by: TechPowerUp
What broke:
  • GPU fan detection malfunction — only one fan spinning
  • Clock stability issues
  • General driver instability
  • NVIDIA action:
    • Driver pulled from download pages
    • Users told to roll back to 591.86 WHQL

2. GeForce 595.71 WHQL — Replacement Driver Still Causing Problems

Issue type: Voltage cap + overclocking failures
Reported by: HotHardware, Neowin, TechPowerUp
What broke:
  • Overclocked GPUs had voltage capped, preventing expected boost clocks
  • New visual artifacts in Resident Evil Requiem
  • Game crashes in Star Citizen
  • Browser DRM playback causing driver timeouts
  • NVIDIA action:
    • Released 595.76 Hotfix to address these issues
    • 595.76 Windows‑update conflict — Users reproduced:
      • Stuttering
      • Black‑screen events
      • General instability
      • The regression was tied to January/February Windows 11 cumulative updates interacting with the driver. Issues disappeared when rolling back Windows updates.

3. GeForce 596.xx Hotfix Thread — Ongoing Regression Reports

Issue type: Resolution scaling & control panel regressions
Reported by: NVIDIA Forums
What broke:
  • Custom resolutions not applying
  • Scaling options refusing to change
  • Regression traced back to post‑591.86 drivers
  • NVIDIA action:
    • Hotfix 596.02 released to address targeted issues
 
Back
Top Bottom