Discussion RDNA4 + CDNA3 Architectures Thread

[DisEnchantment]

With the GFX940 patches in full swing since first week of March, it is looking like MI300 is not far in the distant future!
Usually AMD takes around 3Qs to get the support in LLVM and amdgpu. Lately, since RDNA2 the window they push to add support for new devices is much reduced to prevent leaks.
But looking at the flurry of code in LLVM, it is a lot of commits. Maybe because US Govt is starting to prepare the SW environment for El Capitan (Maybe to avoid slow bring up situation like Frontier for example)

See here for the GFX940 specific commits

History for llvm/lib/Target/AMDGPU - llvm/llvm-project

The LLVM Project is a collection of modular and reusable compiler and toolchain technologies. - History for llvm/lib/Target/AMDGPU - llvm/llvm-project

github.com

Or Phoronix

More AMD "GFX940" Enablement Work Landing In LLVM - Phoronix

www.phoronix.com

There is a lot more if you know whom to follow in LLVM review chains (before getting merged to github), but I am not going to link AMD employees.

I am starting to think MI300 will launch around the same time like Hopper probably only a couple of months later!
Although I believe Hopper had problems not having a host CPU capable of doing PCIe 5 in the very near future therefore it might have gotten pushed back a bit until SPR and Genoa arrives later in 2022.
If PVC slips again I believe MI300 could launch before it

This is nuts, MI100/200/300 cadence is impressive.



Previous thread on CDNA2 and RDNA3 here

Question - Speculation: RDNA3 + CDNA2 Architectures Thread

Man I have been dying to make this one for a while now. First rumours for RDNA3 are here so new thread time! Just going to start off with this one for now: kopite7kimi on Twitter: "@VideoCardz Ah, I mean a simple mcm design with 10240 cores is not enough. Because the lift from RDNA2 to RDNA3...

forums.anandtech.com

 

[blackangus]

Framegen on top of low FPS also creates more discernible artifacts, not to mention the fact that it doesn't work very well with adaptive sync so stuttering gets more noticeable.
I have a 165Hz monitor but I don't really use framegen even on the RTX5090. The setbacks are a bit too great for me, IMO. I prefer cranking up IQ until I reach 80-90FPS.

Im surprised.
The extra motion clarity that can be seen at 240 hz is a huge plus over 80-90FPS. (Im assuming the same at 165 vs say 90)
When I play something at that FPS I cant really tell any input lag differences if I add 2x or 4x framegen on top of that.
While there used to be alot of artifacting, with the current DLSS (unsure about FSR4) I dont really notice any artifacting at all.
I dont have any sync/stutter problems with the MSI I have, this was the biggest benefit to me, at 240hz framegen there is nearly no stutter for anything I play.

If your base FPS is 80-90, what trade-offs are you seeing? (Just the adaptive sync?)

 

[basix]

Maybe something like GSync Pulsar could be a nice motion clarity uplift and then FG has less of an impact on that. But a long as that "Pulsar" is not available for OLED (which is technologically viable to do), I will not change my monitor.

 

[ToTTenTranz]

If your base FPS is 80-90, what trade-offs are you seeing? (Just the adaptive sync?)

Stuttering - because GSync ceases to work - and artifacts, like the objects that appear and disappear beyond the main character.

 

[blackangus]

Stuttering - because GSync ceases to work - and artifacts, like the objects that appear and disappear beyond the main character.

Interesting the different experience, as I have nearly no stuttering at all. (When I say nearly I mean its game induced vs display induced)
I get the most fluid experience I have had since a CRT decades ago.

I havent seen objects appear and disappearing since FSR2. Maybe it happens but it so rare I never notice anything. (I did with FSR2 and some games, and didnt use early DLSS as I had AMD cards previously)

 

[pj-]

Stuttering - because GSync ceases to work - and artifacts, like the objects that appear and disappear beyond the main character.

I don't have an nvidia card but the internet says frame gen is expected to work with gsync so something may be off in your settings

I was a frame gen sceptic but I've been using fsr3.1 fg on my 7900xtx in Stalker 2 for about 30 hours and am quite happy with the performance. Minimal latency, native res at 3440x1440, mixture of high/epic settings, no stuttering except very rare instances that feel like unreal engine being unreal engine, and around 120-140 fps most of the time.

Definitely a lot of visual artifacts from being fsr3.1, but the smoothness of frame rate more than makes up for it.