Ha, also realized I never came back and updated this thread to note that the original goal has been achieved and we can indeed run MAX models on AMD RDNA GPUs today. I’ve also been hacking on some enhancements to matmul and 2-D convolution for RDNA 3+ GPUs that I mention above, which have significantly improved performance over our initial naive implementations of those kernels. Models like FLUX.2-klein actually run fairly well locally on an AMD Strix Halo system (Framework Desktop) using MAX in our latest nightlies.
BradLarson
(Brad Larson)
9
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Initial support for running MAX models on AMD RDNA GPUs | 0 | 97 | February 17, 2026 | |
| Support for Turing Architecture? | 9 | 307 | May 4, 2025 | |
| Modular 25.4: One Container, AMD and NVIDIA GPUs, No Lock-In | 0 | 82 | June 18, 2025 | |
| Examples of custom CPU / GPU operations in Mojo | 29 | 1472 | October 6, 2025 | |
| 🚨 New video: Modular now runs on AMD MI300X and MI325 GPUs | 0 | 91 | June 11, 2025 |