Friday, August 7th 2026
AMD RDNA 4m iGPU Support Arrives in Open-Source Mesa GPU Driver
AMD's mysterious RDNA 4m integrated graphics have reappeared after a few months of silence, as the GPU has officially been merged into the Mesa 26.3 open-source driver stack. Running under the GFX1171 GPU target, this RDNA 4m is neither a true RDNA 4 design nor a simple RDNA 3 GPU IP. It sits just above RDNA 3.5, with supposed capabilities to support INT8 data types and native FSR 4 technology. However, since AMD announced that older RDNA generations are now getting FSR 4 upscaling, the true purpose of RDNA 4m remains unclear.
As readers may recall, AMD plans to keep its RDNA 3.5 iGPU in use until 2029, meaning most consumers buying laptops in the coming years will experience the same integrated graphics capabilities as those found in today's Ryzen AI 300 and 400 series processors. This suggests that the RDNA 4m, despite its limitations, will be a welcome addition to integrated graphics. "Redstone" features such as ML-based upscaling and rendering technologies rely on the RDNA 4 IP to function effectively. With RDNA 4m support now merged into the Mesa open-source graphics driver, the RADV and RadeonSI drivers are being updated for RDNA 4m.We can expect the first appearance of RDNA 4m in "Medusa Point," which will introduce a desktop-exclusive RDNA 4 with the new RDNA 4m variant. In contrast, "Medusa Halo" will utilize AMD's next-generation RDNA 5/UDNA GPU microarchitecture. Although initial comparisons between the two are lacking, some instruction set extensions, such as WMMA and SWMMAC instructions, indicate support in the new "GFX1170" GPU, which should be associated with the GFX11 generation, also known as RDNA 3. Currently, this is believed to be an upgraded RDNA 3 with many RDNA 4 modules, enabling FSR 4 support even on the less powerful "Medusa Point" APU.
Source:
Phoronix
As readers may recall, AMD plans to keep its RDNA 3.5 iGPU in use until 2029, meaning most consumers buying laptops in the coming years will experience the same integrated graphics capabilities as those found in today's Ryzen AI 300 and 400 series processors. This suggests that the RDNA 4m, despite its limitations, will be a welcome addition to integrated graphics. "Redstone" features such as ML-based upscaling and rendering technologies rely on the RDNA 4 IP to function effectively. With RDNA 4m support now merged into the Mesa open-source graphics driver, the RADV and RadeonSI drivers are being updated for RDNA 4m.We can expect the first appearance of RDNA 4m in "Medusa Point," which will introduce a desktop-exclusive RDNA 4 with the new RDNA 4m variant. In contrast, "Medusa Halo" will utilize AMD's next-generation RDNA 5/UDNA GPU microarchitecture. Although initial comparisons between the two are lacking, some instruction set extensions, such as WMMA and SWMMAC instructions, indicate support in the new "GFX1170" GPU, which should be associated with the GFX11 generation, also known as RDNA 3. Currently, this is believed to be an upgraded RDNA 3 with many RDNA 4 modules, enabling FSR 4 support even on the less powerful "Medusa Point" APU.

32 Comments on AMD RDNA 4m iGPU Support Arrives in Open-Source Mesa GPU Driver
We were not talking about if mobile was a good business decision or not. We were taking about how AMD is already selling mobile CPUs, so they should put the effort in to promote them. Why you sidetracked to datacenters is beyond me.Not really. FSR4 has not been confirmed for rDNA 3.5 from AMD. It HAS been cofirmed from 3, but 3.5 is still unknown.
www.techpowerup.com/349718/amd-no-definitive-decision-on-fsr-4-1-support-for-rdna-3-5-apusAnother strawman. We are not talking about fighing the HDMI consortium, we are talking about AMD supporting software upscaling on older architectures and the relevance of rDNA4m.Irrelevant. We are talking iGPUs here, not CPUs.Again, what does intel's dGPU business have to do with them using their newest uARCH in their products?
Their cheaper low end models ALSO use the newest arch, so your bellyaching about top iGPU laptops and handhelds means nothing.
The deal with nvidia has also not stopped intel from using their newest arch.
So no, this is not "give or take", and market segments have nothing to do with this. You're wandering all over the place and not addressing the main point, which is that nothing is stopping AMD from using newer rDNA arches for their iGPUs. Likely because you dont actually have an argument that makes sense, so we have to start talking about nvidia deals and high end laptop prices.
It's might be beyond you because you wish to remain wilfully blind as to where more effort, time, money and business decisions are oriented towards. They can't be everywhere to the same extent and I don't think they want or need to be everywhere at the same time. It's enough that they are competitive and successful where it matters, which they are, more than at any time in history. Unfortunately, gaming community is slowly realising that they are not in the centre of global business attention. Tough.That's correct, but they never promised it might happen. We know that DLL injection files work on 3.5 APUs, so they might need more time to work on it before announcing anything publicly. They first need to optimise INT8 upscaling models for older discrete graphics, and then look at other segments, if viable. Again, things are delivered one by one. I would not make any public promises if I was not sure how well the INT8 model could work on diverse iGPUs. There's a lot to address and test here, anything from 2 CUs all the way to 40 CUs. That's a huge span and meaningful performance uplift may not land everywhere. We also know that performance hit is 6-9% on average on 7800XT, as recently tested by ComputerBase, so they need to decide whether it's worthwhile to roll it out everywhere by optimizing and testing all possible engines and games. Stay tuned. It might arrive to 3.5, or it might not.FSR4 works pretty well on RDNA3 cards already, but benefits have been somewhat limited to Windows environment. Also, you could try to connect a few dots. The upscaling feature is rather pointless in Linux enviroment if you don't have all components of the ecosystem enabled. And HDMI 2.1 is one of those components that can unlock upscaled images in many games beyond 1440/60Hz and 4K/60 10-bit on extrenal displays, such as large home TV. That's one of reasons why Valve folks, AMD engineers and independent developers fought to enable those features in open source drivers, so that users of different GPU engines that do not run on Windows could also benefit the upscaling in HDMI 2.1 environment. By brushing over it, you are underestrimating the importance of fully enabled non-Windows environments. It took more than one year of working on software patches to enable this. It does not grow on trees.iGPU, as the name reads, is a part of CPU. By 2029, there will be plenty of different iGPUs in different devices based on RDNA4m, RDNA5 and likely its refresh or new architecture. It will be a mix of different APUs in the market, as usual. When I bought a brand new laptop with RDNA2 iGPU in 2022, the mainstream selling devices were still APUs from Cezanne and Renoir iGPU, as this was what majority of people were buying. The market has space for several architectures at the same time. Saying that "RDNA4m will be old by 2029" is completely meaningless. Intel is selling right now, in mobility segment, chips from Raptor Lake, Meteor Lake, Lunar Lake, Arrow Lake and Panther Lake. Buyers around the world buy all of those in different devices. How many iGPU architectures is that? Go figure.It's pretty simple. They will use iGPU architectures as they find fit. If they are not competitive, they will lose some market share. If they are competitive and diverse, they will gain. RDNA4m will be for mainstream market, within Medusa Point APU. This one does not need to be particularly powerful. It's for everyday laptops. A CCD could be attached to it for laptops that will feature discrete GPU. More powerful iGPU will feature in Medusa Halo. It is said to be a RDNA5 chiplet, AT3 or AT4. We will see how this one pans out. RDNA5 will also feature in iGPU of new consoles. This will be powerful for console gaming. Plenty of new stuff coming in 2027 and later.
After those we will have Zen 7 based Grimlock Point / Halo, but we don't even have rumours for whatever GPUs will be used on those.
What we may actually see is the Medusa Halo based discrete GPUs, if the rumours are true.
Try this one
AMD does not support end users with their laptops but delivers parts for those solely. Laptop vendors do laptop support not AMD :)