I saw a report of lower Forza Horizon 6 performance on Nvidia driver 610.43.02 compared with 595.71.05 on Linux. What should I look at when comparing the two?
Has anyone noticed worse Forza Horizon performance with newer Nvidia drivers?
Answers
The reported benchmark drop was around 9% on the extreme preset, both with TAA and with DLAA preset M. That is worth checking, but it is one comparison on one setup, not proof that every Nvidia system will behave the same way.
For a useful comparison, keep the game settings and compatibility layer version the same between driver tests. Run the same benchmark more than once and compare frame rates and frametimes, not just a single average.
Reply to this answer
Don't overlook the tradeoff. The newer driver reportedly had more consistent frametimes and felt smoother, even though the benchmark frame rate dropped and the stutter count didn't reflect that impression. Lower average performance and smoother delivery can happen together.
Reply to this answer
There is also a compatibility difference in the report. With 610.43.02, the game reportedly launched without custom VKD3D options. On 595.71.05, it crashed without those options even with the tested Proton version.
So if you're trying the older driver, keep that launch configuration in mind. A faster benchmark isn't much help if the game won't start.
Reply to this answer
High GPU utilization doesn't settle the question by itself. The Forza test reported 98% usage but lower power draw than the GPU's stated maximum. Separately, another report described low and fluctuating utilization in DirectX 12 games on 610 and newer.
Those are different symptoms, so I wouldn't assume they have the same cause. It is also worth testing another game. The Forza comparison included a quick Cyberpunk 2077 run that did not show a regression.
Reply to this answer
Add your answer
Answers are reviewed before publication.
Related Discussions