In an interview with TheStreet, AMD Rick Bergman suggested that RDNA3 architecture should provide a similar performance per watt improvement over RDNA2.
AMD talks RDNA2 ray-tracing and super-resolution
AMD’s Eric Bergman, the former CEO for Synaptics now working for AMD as Executive Vide Present has been asked in the interview about the future of Radeon graphics, including the upcoming Radeon RX 6000 series and its successor.
Unfortunately, AMD did not yet disclose too many details about the ray tracing performance of its RDNA2 based on NAVI 21 GPU, the heart of Radeon RX 6900, and the 6800 series. Although the manufacturer made a commitment it will support the industry standards such as the upcoming Vulkan Ray Tracing API and Microsoft DirectX Ray Tracing.
Next week, on November 18th first reviews of the AMD Radeon RX 6800 series will be published by independent tech journalists. We have heard that they are already in possession of the working drivers, hence we might learn about the true ray tracing acceleration potential of the RDNA2 architecture very soon. That’s in case AMD will not disclose any details themselves.
AMD Navi 21 GPU with Ray Accelerators
In the interview with TheStreet Eric Bergman revealed that AMD’s performance goal is 1440p resolution:
Eric Bergman: And our goal was at 1440p [resolution], to have a great ray-tracing experience. And that was kind of the performance level that we targeted. Now it depends on particular games and everybody’s systems and so on, but I think you’ll find that we have very good ray-tracing performance overall. And the game support will be strong as we go through 2021, because again, we get that great leverage. It’s just built in: You support ray tracing on Microsoft or Sony [consoles], you’re supporting AMD on the PC side as well.
Of course, this was the initial (as of now official) target. We do not know if the statement refers to the upcoming Radeon RX 6900 XT graphics card as well (featuring the full-fat Big Navi).
We have also recently covered a topic of AMD’s potential answer to NVIDIA DLSS technology (deep learning super sampling – AI super-resolution upscaling). AMD has not confirmed when will such technology be available, but Bergman confirmed that for now it is called FSR (FidelityX Super Resolution) and that feature should be implemented eventually:
Eric Bergman: We don’t have a lot of details that we want to talk about. So we called [our solution] FSR — FidelityFX Super Resolution. But we are committed to getting that feature implemented, and we’re working with ISVs at this point. I’ll just say AMD’s approach on these types of technologies is to make sure we have broad platform support, and not require proprietary solutions [to be supported by] the ISVs. And that’s the approach that we’re taking. So as we go through next year, you’ll get a lot more details on it.
Bergman was also asked about the future GPU architecture – RDNA3 – performance per watt improvements, whether similar architectural gains are to be expected when moving from RDNA2. He confirmed that AMD has the same (aggressive) commitment to such performance per watt improvements for RDNA3:
Eric Bergman: Let’s step back and talk about the benefits of both. So why did we target, pretty aggressively, performance per watt [improvements for] our RDNA 2 [GPUs]. And then yes, we have the same commitment on RDNA 3.
[…] If you can improve your perf-per-watt substantially. On the notebook side, that’s of course even more obvious, because you’re in a very constrained space, you can just bring more performance to that platform again without some exotic cooling solutions…We focused on that on RDNA 2. It’s a big focus on RDNA 3 as well.
AMD RDNA3 architecture with Navi 3X GPUs
Another interesting part of the interview is the question about Infinity Cache on RDNA3 GPUs. Bergman thinks that if the RDNA2 Infinity Cache brings performance benefits, then RDNA3 Infinity Cache will definitely be considered as an option:
Eric Bergman: We want to look forward and see what architecture will scale going forward. [With] Infinity Cache, the performance benefits, the performance-per-watt benefits, the cost benefits [made it] a pretty easy decision to make….I don’t want to talk about our next generation [of products], but as you can imagine, when you get those benefits, it’ll certainly be on the table for our next generation.”
AMD has not confirmed the launch year for the RDNA3 series, but according to the official roadmap, the architecture should appear by 2022. The graphics cards based on this architecture will use a new fabrication process (AMD did not confirm the foundry and the node yet). It is possible that AMD will be focusing on delivering the next-gen ‘ultimate platform’, featuring both RDNA3 and Zen4 CPUs.
AMD will now be focusing on its transition to the AM5 socket platform supporting DDR5 memory (and maybe even PCIe 5.0). We do not know if these are Ryzen 6000 series, or if there are any possible architecture refreshes in the channel, but Bergman did say in the same interview that such a possibility is also considered. Please check the link below for the full review.
Rick Bergman is Executive Vice President, Computing and Graphics Business Group at AMD