NVIDIA GeForce 517.48 WHQL driver download & Discussion

Max Perf doesn't work around driver issues, it works around silicon faults and card defects.

 

What do I know what specific things are called aside from shimmering issues.
I can point at quite a few old games where developers did a lousy job regarding LOD, mipmaps or both, were using HQ and clamping the LOD BIAS straightens shimmer out to some extent, but still not where it would be if the same game developers would have done a proper job.
At some point game developers did fix it for a few of those games, not Nvidia.

My mistake to assume such, no idea what is going on there, probably some background crap that fires up.

 

Interesting, so what you're saying is that my brand new card might be possibly faulty ?
*But why is the older driver working just fine without that setting enabled then*

And also why does it only happen when the card is idle, and never during high load

 

clamping lod bias does nothing beyond kepler, need to change some profile fields and use the lod offset.

processing errors when a transistor's switching frequency increases faster then the voltage fed into it.
stuck transistors
native assembly that some part of the chip doesn't handle well at low voltage.

pretty much the same reasons a cpu can experience a silicon fault.

 

I wish there was a way to monitor power states to see how things are different between 517.xx and 472.xx,
the only thing I can see is my card quite often sitting at P8 then increasing to P0, however when forced it's always at P0.

I will check whether that was different with 472.xx driver,- could it be forcing it to P0 all the time, hmm.

 

It seems 472.xx drivers are stable for me though, without forcing it to "Max Perf" ...
Yet the new ones are only stable when forced,- I'm still not 100% convinced it's a faulty card.

I can see it's putting the card at P8 correctly like with the new driver, unless the new one tries to go even lower if that's possible ...

What are your thoughts ?

 

I think my issues have to do with this, never fixed it seems:

• [RTX 30 series] PC monitor may not wake from display sleep when GPU is also connected to an HDMI 2.1 TV, and the TV is powered off. [3645633]

 

526.47 fixed the issues!