Slight and Little are not indicating more than difference between sessions from day to day and has nothing to do with:
[Updated]
I was trying to find some information about the latest Nvidia drivers, from 446.14 to 461.40, and I came across this user doing a lot of benchmarks. I didn’t spend much time into all the details he is publishing but it might be interesting to some users:
@AlexE75 I might have wrongly express myself because when I read your comments I have a feeling you’re thinking I’m trying to arguing with you which I’m not. You rightfully corrected my wrong assertion about the security bulletin, for which I’ve amended the “recommended for VR” post, and I’m not here arguing about anything AGAINST you. I’m just saying there should be little changes for the Pascal architecture for a certain number of driver versions, because most efforts in the drivers are naturally drawn to supporting the new Turing, then Ampere architecture, and there are some commonality in the architecture for which any change benefits all these across the board, but most release notes are insisting on Ampere and Turing centric enhancements and bug corrections mostly for while now. This is why I believe (but only believe) there shouldn’t be as dramatic a change from 461.xx to 461.yy on Pascal, whereas it could be dramatic a change on Turing and Ampere. And this doesn’t preclude there is a change you can measure and notice on your hardware as well, I just think it shouldn’t be expected to show huge changes, which you agree on when qualifying it as “little” yet “noticeable”.