Discussion RDNA 5 / UDNA (CDNA Next) speculation

Page 95 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

luro

Member
Dec 11, 2022
102
142
86
They intend to drop support of rdna2 while those GPUs still on shelf and all the rdna lines that should support ROCm still has serious issues running it

It's actually a lot worse than that.

AMD dropped support for RDNA3/3.5 which is an architecture coming in future products like Gorgon Halo.

At the moment, there's hardly any reason to trust AMD's GPU support at all.
Buyers of recent RDNA4 GPUs are likely to regret their decision in a couple of years when RDNA5 is out there.

Their latest driver literaly has a bunch of fixes for RNDA2 and RDNA3. I dont know how this is not supporting.
 
  • Like
Reactions: Win2012R2 and NCU

ToTTenTranz

Senior member
Feb 4, 2021
962
1,579
136
I don't know about ROCm, but at least the driver thing AMD somewhat backtracked from was "only" about RDNA1+2, and more about Day 1 support + optimizations for new games, rather than total droppage.

RDNA3 is in "maintenance mode" just as much as RDNA1+2. FSR4 and Redstone were exclusive to RDNA4 and there's zero sign of this changing.


Their latest driver literaly has a bunch of fixes for RNDA2 and RDNA3. I dont know how this is not supporting.
Yes, fixes means "maintenance mode". Support should mean game optimization and support for newly launched software features.
 
  • Like
Reactions: tajoh111 and MrMPFR

tajoh111

Senior member
Mar 28, 2005
354
396
136
It's not worse at all.
NV just doesn't talk about it.
Nvidia is definitely supporting it's last gen better than AMD. Not just from a features standpoint but optimizing games too.


Nvidia has been gaining ground on AMD last gen vs last gen.

The RTX 4080 is now faster than a 7900xtx and the gap between a 7900 xtx vs 4090 is nearly 39. percent(where it used to be 29%). The 4070 regular is now beating a 7800xt where it previously lost.


The RX 7600 is 31% slower than a rtx 3060 ti where it used to be 17% slower.

With new releases aside from rare exceptions like call of duty, AMD cards tend to underperform in regards to their last gen while Nvidia cards perform at normal expectations. The 6700 Xt performance is underwhelming too. Its slower than a rtx 3060 ti even with 8gb of vram. Bring in AMD ultra underwhelming support and launch of redstone, lack of support of FSR4 on anything aside from RDNA 4 and it just confirms what you already know. AMD does not care about client GPU.
 

gdansk

Diamond Member
Feb 8, 2011
4,796
8,097
136
The RTX 4080 is now faster than a 7900xtx and the gap between a 7900 xtx vs 4090 is nearly 39. percent(where it used to be 29%). The 4070 regular is now beating a 7800xt where it previously lost.
The test suite changed. Even something as simple as using a different, faster CPU will favor Nvidia's CPU heavy drivers.

And other test suites disagree with that assessment.

Go spread your multivariable schlockbenches/FUD in another (non-vendor) thread. RDNA3.x isn't in maintenance mode despite not having all features. Much like Ampere isn't in maintenance mode despite not having MFG.
 
Last edited:

Win2012R2

Golden Member
Dec 5, 2024
1,328
1,368
96
Yes, fixes means "maintenance mode". Support should mean game optimization and support for newly launched software features.
It's unrealistic to expect any of the following:
1) new software features like FSR4 that rely on new hardware added
2) doing anything other than maintenance mode for product lined that EOL-ed
3) doing game optimisation in drivers - drivers should have never done that, it's the job for devs to optimise their games that they sell: driver should just be driver that complies with the spec and does it as fast as possible.
 
  • Like
Reactions: marees

tajoh111

Senior member
Mar 28, 2005
354
396
136
The test suite changed. Even something as simple as using a different, faster CPU will favor Nvidia's CPU heavy drivers.

And other test suites disagree with that assessment.

Go be spread your multivariable schlockbenches/FUD in another (non-vendor) thread. RDNA3.x isn't in maintenance mode despite not having all features. Much like Ampere isn't in maintenance mode despite not having MFG.
If changing the test suite alter the results in favor of Nvidia, it proves my point correct. That is AMD results with older games or games released during its retail life show better results as a result of AMD optimizing for that generation. If AMD last gen is losing ground with new games vs Lovelace and RDNA4, it means AMD is not optimizing for newly released for its last gen.

E.g look at Anno 117 results below. These results do not include ray tracing. Lovelace performs where it should. RDNA 3 on the other hand does not and RDNA 4 towers over it.

PAX117.jpg

Same thing but to a lesser extent happens in Battlefield 6.

battlefield6.jpg

Looking at the 7900xtx and 7800xt results compared to the the 9700xt. Now compared Blackwell vs Lovelace. Nvidia cards are performing where they should while AMD's last gen is not.

Adding new games to test results and last gens card performing well in them show if a vendor is optimizing for these new games for their last gen(i.e optimizing rather than just bare minimum support). AMD fine wine was derived from this. Remember this used to be the case of the 7970 or RX 580 for example. These cards aged much better than the gtx 680 or gtx 1080.

It is not just computerbase.de that is show this regression, it is techpowerup too.


Note these results also are at 4k(atleast with the rtx 4080, 4090 vs 7900xtx) thus the CPU affect should be negligible. AMD should be gaining ground even more than before at 4k because of their vram advantage. If they are not and are losing ground, it means AMD is just doing the bare minimum to ensure compatibility and stability. Nvidia rtx 4070 should not be gaining ground at 4k with a 12gb vram buffer vs a 7800xt.
 

gdansk

Diamond Member
Feb 8, 2011
4,796
8,097
136
it means AMD is just doing the bare minimum to ensure compatibility and stability
No, it doesn't. Your multivariable tests show nothing of the sort.

You can easily look at AMD commits to llvm forks and linux in the last year to show it isn't in maintenance mode. They had made changes to their compilers that we can see in public for older architectures. Your worthless theory crafting ignoring architectural differences, testing setup variation and game selection is falsified easily by looking at real, shipping code which improves performance on e.g. gfx11.
 
Last edited:
  • Like
Reactions: NCU