Well then there’s VR performances, nvenc, DLSS, frame gen, superior reflex latencies. Resell value of CUDA and Machine learning for any professionals
I leave RT on all the time
plus the cool path tracing projects.
What would I get for say, picking a 7900XTX? ~2-4% higher rasterization performances with higher power consumptions? Only to lock my monitor to 120 fps and waste more performances? The rasterization race is getting meaningless, these games perform well when it’s not a shit port. Games that stress hardware in sub 60 fps range are RT games and DLSS brings them back higher.
I don’t get why honestly.
And I was ATI-AMD for 20 years, from their Mach series up to 2016. It used to be just about fps vs $ vs power. ATI was very aggressive on price and I kept buying them. Things are a lot more complex now for decisions. I also went through the tougher driver periods of ATI/AMD.. I know RDNA 2 & 3 are pretty good now, but I also had my brother with a 5700XT with constant weird problems and black screens until he sold it. That was not that long ago. AMD subreddit had like 2 years worth of peoples having troubles with these cards.
RDNA 2 I even attempted to get because getting
any GPU during that period was a win, but if you put a 6800XT and 3080 on the table in 2020 for $50 difference? Why would I go AMD?
I honestly don’t know what tour de force they have to do to gain massive market share back. Pity buying is not a solution.