Chazore said:
Sounds like it's gonna be expensive and not that great at the same time. Doesn't bode well tbh. Guess I will be sticking with Green team for another round if this keeps up. |
hinch said:
I like to call it early adopter tax - ala like beta testing. Okay its a bit harsh calling it that lol. Yeah I would like to see more Path Tracing. That can make an average game look amazing. Even for games like Quake II and Minecraft the results are staggering. It really does give a massive difference in visual fidelity and opens up other opportunities in gameplay.
Yeah, its AMD's first attempt with RT and still no real competition for the machine learning of Nvidia (DLSS) - still not convinced that Direct ML can give similar results to latter. And still lagging years behind Nvidia. RDNA 2 might have been competitive last year but with Ampere coming its going get steamrolled. Big die, will run hot and probably very expensive. Unless they price it much lower (can't see it happening) it looks to be just a slightly better Turing in compute and maybe worse performing in RT. Hope I'm wrong but eh we'll see. |
Yea we will see how it goes but I am starting to believe those rumours about Big Navi being 3070/3070 Ti performance.
The one thing to remember is that RDNA 2 does now have full support for Direct X 12 Ultimate while RDNA 1 doesn't. So it should perform better than RDNA 1 on the games that use those Direct X 12 Ultimate feature set so there are changes for sure. The kicker is Turing's RTX GPUs also has full support for Direct X 12 Ultimate.
So in theory, for those games that don't use Direct X 12 Ultimate, it could be that RDNA 2 performs similarly to RDNA 1. But those that use Direct X 12 Ultimate, RDNA 2 should see performance gains.
But when Ampere arrives, both AMD and our wallets still start to sweat.
PC Specs: CPU: 7800X3D || GPU: Strix 4090 || RAM: 32GB DDR5 6000 || Main SSD: WD 2TB SN850