Actually I have both a 6900xt and a 3090 both liquid cooled both the same system exactly bar the GPU, I also feel this Benchmark is unfair as it targets specific features of the Nvidia cards, on a Day to Day basis if we take away the specific RTX requirements from games and just base it on performance of the GPU to render frames then both look the same, the FPS is pretty identical within a few percent and to be truthful the 6900xt uses a lot less power to render the same images and frame rates.
Singling out specific features as this does is just painting yourself into a corner or being paid handsomely to make a certain party look very good and its technology very relevant,
Actually I have both a 6900xt and a 3090 both liquid cooled both the same system exactly bar the GPU, I also feel this Benchmark is unfair as it targets specific features of the Nvidia cards, on a Day to Day basis if we take away the specific RTX requirements from games and just base it on performance of the GPU to render frames then both look the same, the FPS is pretty identical within a few percent and to be truthful the 6900xt uses a lot less power to render the same images and frame rates.
Singling out specific features as this does is just painting yourself into a corner or being paid handsomely to make a certain party look very good and its technology very relevant,
It's simple:
Please, could you say the DX12/Vulkan instructions that AMD doesn't support?
AMD affirms their RX 6000 are DX12 Ultimate cards
https://www.amd.com/en/technologies/directx12
If it's true, the difference is at hardware level, and that is AMD fault, not a benchmark's biases.
If you want make a benchmark, you must crush the hardware, all the hardware.
If the future RX7900 destroys the RTX 4090 in this test, will it mean that the test is made to favor AMD?
No, it would mean that AMD has improved a lot in the parts that this test uses.
The same thing happened years ago with tessellation, when ATI users complained that games used it too much, but forgot about it as soon as ATI/AMD cards got good enough.
outside the fact vulkan is another extension crapfield by khronos group, you can write code in ANY API that gives huge advantages to a specific hardware instead to another. It always has been.
rubbish benchmark for nvidia oriented effects should be used to test nvidia only cards
Undying:
Even the port royal isnt so punishing for amd cards. It makes no sense using something as a regular benchmark that favors one side so much.
D3D12 and Vulkan Raytraced performance.
That is where those AMD cards got their weakness, without they are pretty much on pair, better in some games and worse in others.
even as a 3090 owner I'm thinking what's the point of this thing we already know AMD can't accelerate RT
RT that I barely use and which sometimes give a garbage result example : metro exodus rt edition that looks like 360p wth
forza looked nice but so much nice than without rt ? not really it's great for screenshots
Yes, it is absolutely believeable that a 3070 is faster than a 6900 XT while in games the 6900 XT is on par with the RTX 3080 - or a bit faster.
and games prove that too,at least the more rt-demanding ones with full resolution raytracing
this isn't too bad for amd tbh,check 3dmark full path tracing on guru3d reviews,they totally crumble.
Looks like a decent RT benchmark. However, I feel like the results just confirm what we already knew; that the top AMD cards have RTX2080ti-like RT performance.
[spoiler]
[/spoiler] [spoiler]
[/spoiler]