Honestly at least the "exclusivity reasoning" that Nvidia gives sort of makes sense. There are games that let you run Ray Tracing on Pascal/RDNA 1 and they really do cuck the performance on both very hard. You could technically run DLSS on Shader Cores similar to how DLSS 1.9 was instead of Tensor Cores but the issue is that by doing so, you have too many things trying to fight for the same resource... Where as if you just off load DLSS to Tensor cores, the shader cores can focus more on producing frames instead of producing frames and computing upscaling. DLSS 3 is similar in that fashion. A 3090 for example can do 285 Tensor-TOPs vs a 4060, that's right a 4060 can do 242 Tensor-TOPs. A 4090 can do 1321 Tensor-TOPs.
So to me, on paper, it at least makes sense even if it's lame.
The problem with Radeon exclusivity is that they are making Anti-Lag+ Exclusive and Hypr-RX exclusive because RDNA 3 architecture can apparently do something that RDNA 2 and older can't. But in the case of those two features it's like huh? moment. Hypr-RX is just a one click toggle that just enables 3 Radeon features which you can customize. Why does that need to be RDNA 3 exclusive since RDNA 2 can do Boost, Anti-Lag and FSR/FG? And Nvidia has had Reflex which is similar to Anti-Lag+ since Maxwell. Are you telling me that Nvidia had the foresight to implement a hardware feature 9 years ahead of Radeon? It just doesn't sound right but even if that is the case, I still don't see why Hypr-RX is exclusive to RDNA 3.
PC Specs: CPU: 7800X3D || GPU: Strix 4090 || RAM: 32GB DDR5 6000 || Main SSD: WD 2TB SN850
