Official AMD ROCm™ Support Arrives on Windows for ComfyUI Desktop
Seamless ROCm™ integration brings high-performance AMD acceleration to Windows
We’re excited to announce that official AMD ROCm™ support is now available on the Windows ComfyUI Desktop app, starting with version v0.7.0.
With this release, Windows users can finally tap into the full power of supported AMD Radeon™ GPUs and Ryzen™ AI processors, unlocking faster performance and a smoother experience across ComfyUI workflows.
Going forward, users will continuously be able to run ComfyUI on AMD ROCm™ GPUs with:
How to get started:
Download ComfyUI desktop
During installation, AMD ROCm™ should be automatically selected
Finish installation and boot ComfyUI!
This release of the ComfyUI Desktop app and Portable is based on ROCm 7.1.1. Additional performance, feature improvements, expanded support, and bug fixes are expected in future releases.
The AMD ROCm™ 7.1.1 Preview driver is recommended for this release and supports the use of --use-pytorch-cross-attention flag. This support is expected to be merged into the mainline branch in an upcoming AMD Software: Adrenalin™ Edition driver release. Please visit the AMD blog for more details on the benchmark.
In addition to desktop, AMD ROCm™ is also supported on the portable and git versions of ComfyUI.
Happy building!



Huge win for AMD users who've been stuck with suboptimal performance! The automatic ROCm selection during installtion is a nice touch since manually configuring GPU backends can be a pain. I'm curious how this compares to DirectML perfomance on the same hardware, since that was the previous workaround. The cross-attention flag support in 7.1.1 should help a lot with memory-intensive workflows.
A huge thank you for providing such a great tool!
I’ve been running some initial tests on my system, and it’s working very smoothly and perfectly stable with Python 3.12.11.
As a side note : I noticed a significant performance boost when running it on my manual ComfyUI setup with Python 3.13.9. I don't know if it is just the Python version.
in seconds
Py3.13.9 Py3.12.11 #1...Py3.12.11 #2
1........89,43.........317,63.............176,56
2........87,9...........211,21............223,56
3........88,53..........201,74.............318,05
4........84,86.........189,76.............219,25
AVG..87,6975.....230,085...........239,39
%.......100,00 .....262,36.............272,97 %
system:
Win 11
AMD Ryzen 5 7500F 6-Core Processor
AMD Radeon RX 7800 XT
64GB RAM.
test workflow:
https://openart.ai/workflows/aura_111_111/fantasy-elf-sdxl-cinematic-nsfwsfw-portrait-workflow-hires-fix-vae-ksampler-workflow/JiDA2LC9sKQnvBUbNFRQ