All testing was done with Reflex On, but not Boost FYI. I also retested without Reflex (other than the "latency only" option) on the 4060 Ti to check things out. Here are the results (1 = 1080p medium, 2 = 1080p very high, 3 = 1440p very high, 4 = 4K very high, 5 = 4K high, 6 = 4K high + quality upscaling, 7 = 4K high + quality upscale + framegen):
Code:
HorizonForbiddenWest RTX 4060 Ti Reflex Off (1) - AVG: 133.7 99pMIN: 100.3
HorizonForbiddenWest RTX 4060 Ti Reflex Off (2) - AVG: 76.7 99pMIN: 35.9
HorizonForbiddenWest RTX 4060 Ti Reflex Off (3) - AVG: 62.5 99pMIN: 27.9
HorizonForbiddenWest RTX 4060 Ti Reflex Off (4) - AVG: 34.3 99pMIN: 20.2
HorizonForbiddenWest RTX 4060 Ti Reflex Off (5) - AVG: 43.4 99pMIN: 33.3
HorizonForbiddenWest RTX 4060 Ti Reflex Off (6) - AVG: 52.1 99pMIN: 31.4
HorizonForbiddenWest RTX 4060 Ti Reflex On (7) - AVG: 50.2 99pMIN: 36.9
HorizonForbiddenWest RTX 4060 Ti Reflex On (1) - AVG: 129.9 99pMIN: 95.1
HorizonForbiddenWest RTX 4060 Ti Reflex On (2) - AVG: 73.8 99pMIN: 36.2
HorizonForbiddenWest RTX 4060 Ti Reflex On (3) - AVG: 54.1 99pMIN: 27.4
HorizonForbiddenWest RTX 4060 Ti Reflex On (4) - AVG: 34.1 99pMIN: 21.3
HorizonForbiddenWest RTX 4060 Ti Reflex On (5) - AVG: 44.0 99pMIN: 28.3
HorizonForbiddenWest RTX 4060 Ti Reflex On (6) - AVG: 54.6 99pMIN: 43.4
HorizonForbiddenWest RTX 4060 Ti Reflex On (7) - AVG: 48.8 99pMIN: 36.7
If you look at the numbers, clearly the 1% lows ("99pMin") start to show some fluctuations. FrameGen requires Reflex be enabled, and I had a larger than normal delta there as well. GPUs with 12GB or even 10GB seemed to be far more consistent in performance. The 8GB cards on the other hand could get a bit "wonky" at times, and even restarting the game didn't always seem to help.