By using this site, you agree to our Privacy Policy and our Terms of Use. Close
Jizz_Beard_thePirate on 22 August 2020
Edit Thread
Trumpstyle said:
Captain_Yuri said:

Sorry how does everything point towards AMD winning? I'd like them to win but so far, I haven't seen any reason to believe they will.

Probably but every time I hear something new about RDNA 2, it feels more and more like it will be the case. Would like to be proven wrong though, specially my wallet.

And you are basing this on?

The dude that revealed the name Geforce 3090 and Gddr6x, he leaked the exact spec, die size and node for GA100 1 year before release, so we know his info is good. According to him geforce 3090 will pull 350W. Performance I think will be around 35-50% over geforce 2080 ti depending on game and resolution.

For big navi we have 3 clues for performance, first is from AMD themself:

I assume amd is comparing 5700xt to big navi.

Then we have this tweet, he has very similiar info as the previous dude I mentioned, but is not proven yet:

But the big one is simply this leaked benchmark:

Source:https://hothardware.com/news/radeon-big-navi-gpu-leaks-performance-gain-rtx-2080-ti

Showing big Navi beating the highest overclocked geforce 2080 ti by ~20%.

Pull all this togheter I'm guessing big navi will have performance around +35% over geforce 2080 ti and pull about 250-300W. Geforce 3090 will be around 30-50% performance over geforce 2080 ti and pull 350W.

Die size for big navi ~505mm2 and ~627mm2 for geforce 3090.

Edit: Noticed you asked what I based my ray-tracing performance and not big navi vs geforce 3090/3080 ti.

There's nothing reliable I seen for the consoles rtx performance, there was a game developer over at anandtech hinted that PS5 rtx performance is above geforce 2080 ti and there is another user there who said Rdna2 rtx performance is 1.5x above Turing, so I just pulled them togheter and got ~25% above geforce 2080 ti, we shall see if it's correct.

Yea AMD has known to make up nonsense slides based on questionable metrics before... Where are you getting the 35% from?

20% if you are rounding up is rumoured to be at best, on par with a 3080. Not to mention that is a pretty random benchmark instead of something like Time Spy and that benchmark also has some weird results...

So the only benchmark rumour shows less than 20% performance improvement yet you are turning that into 35%?

As for the last Ray Tracing, yea I doubt that's gonna happen. From the hints that we have seen, Neither MS or Sony seem very confident on their hardware to do Ray Tracing very well. If you look at the gameplay that we have seen from PS5, you can see that mainly, it's reflections and on top, it's selective and low resolution reflections where as we have seen the 2080 Ti do reflections at Native 4k.

Of course it's not 1:1 since we won't know until it comes out but if it was faster than Turing, they would certainly be a lot more confident.

Also:

https://www.resetera.com/threads/xbox-series-x-silicon-breakdown-hot-chips-2020-analysis-digital-foundry.272157/page-5#post-43192200

"That is not what we put out, rather I have said that it was close to a 2080 or "broadly similar to" it in the gears 5 benchmark on Ultra with no dynamic res on - in actuality it performed a bit below the 2080 in the gears 5 benchmark as Richard has said in one of his Videos. So nothing about 100 fps or whatever you say there. Regarding RT performance, we have said nothing definitive yet since there is not definitive comparisons out there without Vsync on - so we will have to wait for the game's to Release. But the rumbling in my gut elsewhere does not make me think it will be exactly be RTX 2080 Tiered."

Last edited by Jizz_Beard_thePirate - on 22 August 2020

                  

PC Specs: CPU: 7800X3D || GPU: Strix 4090 || RAM: 32GB DDR5 6000 || Main SSD: WD 2TB SN850