My first impressions on ROCm and Strix Halo

by random_ on 4/18/2026, 9:50:48 PM

https://blog.marcoinacio.com/posts/my-first-impressions-rocm-strix-halo/

Comments

by: IamTC

Nice. Thanks for the writeup. My Strix Halo machine is arriving next week. This is handy and helpful.

4/19/2026, 2:53:15 AM

by: everlier

owning GGUF conversion step is good in sone circumstances, but running in fp16 is below optimal for this hardware due to low-ish bandwidth.It looks like context is set to 32k which is the bare minimum needed for OpenCode with its ~10k initial system prompt. So overall, something like Unsloth's UD q8 XL or q6 XL quants free up a lot of memory and bandwidth moving into the next tier of usefulness.

4/19/2026, 2:44:19 AM

by: JSR_FDED

Perfect. No fluff, just the minimum needed to get things working.

4/19/2026, 2:45:17 AM

by: timmy777

Thanks for sharing. However, this missed being a good writeup due to lack of numbers and data.I'll give a specific example in my feedback, You said:``` so far, so good, I was able to play with PyTorch and run Qwen3.6 on llama.cpp with a large context window ```But there are no numbers, results or output paste. Performance, or timings.Anyone with ram can run these models, it will just be impracticably slow. The halo strix is for a descent performance, so you sharing numbers will be valuable here.Do you mind sharing these? Thanks!

4/19/2026, 1:53:28 AM