AI Inference on AMD Ryzen AI Max Processor (ROCm Blogs)

Fields

id

sources/bench-amd-ryzen-ai-max-uma-ollama

URL

https://rocm.blogs.amd.com/artificial-intelligence/ryzen-uma-llm/README.html

Publisher

AMD

Kind

benchmark

Accessed

2026-10-11

notes

AMD ROCm blog post about AI inference on AMD Ryzen AI Max processor with unified memory and Ollama. Published 2026-05-25 by Chengjia Huang, Daniel Huang, George Wang. Benchmarks Qwen3.5 9B-122B on Ryzen AI Max with 128GB UMA. 2026-10-11: Re-read official AMD ROCm engineering blog post (Chengjia Huang, Daniel Huang, George Wang). Confirmed first-party benchmarks running Qwen3.5 9B-122B on Ryzen AI Max+ with 128GB UMA and Ollama on Ubuntu/ROCm. Raised 0.6 -> 0.80 as verified vendor engineering benchmark.

Notes

AMD ROCm blog post about AI inference on AMD Ryzen AI Max processor with unified memory and Ollama. Published 2026-05-25 by Chengjia Huang, Daniel Huang, George Wang. Benchmarks Qwen3.5 9B-122B on Ryzen AI Max with 128GB UMA.

2026-10-11: Re-read official AMD ROCm engineering blog post (Chengjia Huang, Daniel Huang, George Wang). Confirmed first-party benchmarks running Qwen3.5 9B-122B on Ryzen AI Max+ with 128GB UMA and Ollama on Ubuntu/ROCm. Raised 0.6 -> 0.80 as verified vendor engineering benchmark.

Cross-references

This record references no other records.