Monday, February 20th 2023
AMD RDNA4 Architecture to Build on Features Relevant to Gaming Performance, Doesn't Want to be Baited into an AI Feature Competition with NVIDIA
AMD's next-generation RDNA4 graphics architecture will retain a design-focus on gaming performance, without being drawn into an AI feature-set competition with rival NVIDIA. David Wang, SVP Radeon Technologies Group; and Rick Bergman, EVP of Computing and Graphics Business at AMD; gave an interview to Japanese tech publication 4Gamers, in which they dropped the first hints on the direction which the company's next-generation graphics architecture will take.
While acknowledging NVIDIA's movement in the GPU-accelerated AI space, AMD said that it didn't believe that image processing and performance-upscaling is the best use of the AI-compute resources of the GPU, and that the client segment still hasn't found extensive use of GPU-accelerated AI (or for that matter, even CPU-based AI acceleration). AMD's own image processing tech, FSR, doesn't leverage AI acceleration. Wang said that with the company introducing AI acceleration hardware with its RDNA3 architecture, he hopes that AI is leveraged in improving gameplay—such as procedural world generation, NPCs, bot AI, etc; to add the next level of complexity; rather than spending the hardware resources on image-processing.
AMD also stressed on the need to make the GPU more independent of the CPU in graphics rendering. The company took several steps in this direction over the past many generations, with the most recent being the multi-draw indirect accelerator (MDIA) component introduced with RDNA3. Using this, software can dispatch multiple instanced draw commands that can be issued on the GPU, greatly reducing the CPU-level overhead. RDNA3 is up to 2.3x more efficient at this than RDNA2. Expect more innovations along these lines with RDNA4.
AMD understandably didn't talk anything about the "when," "what," and "how" of RDNA4 as its latest RDNA3 architecture is just off the ground, and awaiting a product ramp through 2023 into the various market-segments spanning from iGPUs to mobile GPUs, and mainstream desktop GPUs. RDNA3 is currently powering the Radeon RX 7900 series high-end graphics cards, and the company's latest 5 nm "Phoenix Point" Ryzen 7000-series mobile processor iGPUs. You can catch the 4Gamer interview in the source link below.
Sources:
4Gamers.net, HotHardware
While acknowledging NVIDIA's movement in the GPU-accelerated AI space, AMD said that it didn't believe that image processing and performance-upscaling is the best use of the AI-compute resources of the GPU, and that the client segment still hasn't found extensive use of GPU-accelerated AI (or for that matter, even CPU-based AI acceleration). AMD's own image processing tech, FSR, doesn't leverage AI acceleration. Wang said that with the company introducing AI acceleration hardware with its RDNA3 architecture, he hopes that AI is leveraged in improving gameplay—such as procedural world generation, NPCs, bot AI, etc; to add the next level of complexity; rather than spending the hardware resources on image-processing.
AMD understandably didn't talk anything about the "when," "what," and "how" of RDNA4 as its latest RDNA3 architecture is just off the ground, and awaiting a product ramp through 2023 into the various market-segments spanning from iGPUs to mobile GPUs, and mainstream desktop GPUs. RDNA3 is currently powering the Radeon RX 7900 series high-end graphics cards, and the company's latest 5 nm "Phoenix Point" Ryzen 7000-series mobile processor iGPUs. You can catch the 4Gamer interview in the source link below.
221 Comments on AMD RDNA4 Architecture to Build on Features Relevant to Gaming Performance, Doesn't Want to be Baited into an AI Feature Competition with NVIDIA
No matter how much you may like AMD, this round fluked in comparison to Navi 21 and Zen 3, which are probably AMD's most solid products in a very, very long time. The extremely poor sales numbers for RDNA 3 lineup in general plus the crazy pack-ins and bundles on Zen 4 show exactly what's happening.
They can either stick to their old guns, or run after innovation, which they cannot afford at the moment. Perhaps, as defeatist as Wang's argument is, it may very well be the best course of action for the company right now, even if the market has clearly spoken that it is interested in Nvidia's technologies and statements: the 4090 is outselling everything else in the market combined right now, even at its absurdly high price. The aforementioned Nokia analogy may very well apply here.
If anything, they have some soul searching to do. Godspeed.
Late edit: I don't hate AMD... Why would I?
With so many graphics people ATi has one has to wonder what they are doing there; or are they peradventure busy patching up all the deficiencies their architecture accumulated through the ages?
Inquiring minds want to know. :)
I don't really see game engines implementing an AI feature that's going to screw over console compatibility, especially for something that you can't turn off like NPC AI.
In my opinion, knowing that CPUs like R5 and R7 are aimed at gamers, the X3D line should have been released from the beginning and the Non X line would be the cheaper alternatives released later. That would make more sense.
Yeah that's exactly what we need AIsplaining :laugh:
Of course it is great if you have more options than less options, but on other hand like mentioned. If you don't use it - then meh.
On 3060 ti have never used tensor cores and ray tracing tried only in minecraft and to get 30fps... cubes, cubes give you 30fps... at that moment I was like. Yeah, I have, but it is not enjoyable so meeh. Not to mention it only works beautiful on nvidia few created maps and that is it.
In the future, when AI performance is important, AMD is much better off adding a Xilinx AI accelerator as a separate chiplet.
Maybe read this conclusion and much like art draw from it what you want to see. Asking another person to speak for someone else is strange.
"While acknowledging NVIDIA's movement in the GPU-accelerated AI space, AMD said that it didn't believe that image processing and performance-upscaling is the best use of the AI-compute resources of the GPU, and that the client segment still hasn't found extensive use of GPU-accelerated AI (or for that matter, even CPU-based AI acceleration). AMD's own image processing tech, FSR, doesn't leverage AI acceleration. Wang said that with the company introducing AI acceleration hardware with its RDNA3 architecture, he hopes that AI is leveraged in improving gameplay—such as procedural world generation, NPCs, bot AI, etc; to add the next level of complexity; rather than spending the hardware resources on image-processing."
For the love of god people read the short article. AMD is not giving up on AI, they just focusing on what will have the biggest impact for gaming on their gaming GPUs.
@Dimitriman AMD announced a DLSS 3.0 competitor some time ago during the launch of RNDA3 GPUs. Technically you are correct, AMD won't have a DLSS3.0 competitor with RDNA4 but that's because they already will have it under RDNA3.
If forming an opinion based on first-hand experience means ignorance to you, then maybe you're the one with the agenda and there's nothing left to talk about.
Stack of cards are Turing gen RTX Quadros, RTX3080/ti/90, RX6700xt/6900xt
FSR2.0 is pretty much on par with DLSS 2.0/2.1 depending on the game.
I have not messed with DLSS 3 as that is frame gen and needs 4k gen that I don't have... I also don't see frame gen as an advantage as I am into competitive low latency play.