Sorry for delay on this, but I wanted to get a few full days of data for comparison.toTOW wrote:Veablo, XanderF > Can we see some logs showing the part where PPD is normal and a part where PPD is reduced ?
Ultimately I also rarely used 511.23, in favor of 511.09 - which was still much worse than 496.49, but not as catastrophically bad.
As above, this exhibits the behavior on 511 of the GPU showing 100% utilization, but much lower temperatures than on 496... 54C compared to 78C.
Curiously, the clock speeds were *higher* on 511 - varying from 1875 to 1900, while on 496 (with the better performance) clock speeds hovered between 1740 and 1815.
EDIT: Something else curious - although not sure how much of an impact this makes - while my card is SKU'd as a "GeForce 3060"...and indeed reports the correct number of cores, clock speeds, etc. It's actually a GA104 (IE., Geforce 3060 Ti/3070) chip - binned down by nVidia to a 3060. This SHOULDN'T affect anything - it should act like a regular 3060 in every way - but worth calling out*.
* I mean, taking a SWAG at a cause - it does mean this chip has no hardware limitations in place to reduce cryptocurrency mining performance, as newer chips have. Rather, nVidia has been limiting crypto performance on this generation of chips via driver changes. Is it possible they goofed with the 511-series drivers, and are nerfing Folding performance as well as crypto performance for this chip?
Logs to follow...
