Q: How fast is qwen3-coder 30B-A3B MoE on the AMD Instinct MI355X 288GB?

~1200–1900 tok/s (active-expert estimate) on the AMD Instinct MI355X 288GB. computed roofline: 8000 GB/s × 0.28–0.45 efficiency window / 1.9GB (Q4_K_M) — bands, never points. MoE: file is 18.6GB (that decides fit) but only ~3B of experts activate per token (~1.9GB streamed) — band below is the active-expert estimate, optimistic; community MoE rows run several× the naive bandwidth math

Question 1

Can the AMD Instinct MI355X 288GB run qwen3-coder 30B-A3B MoE?

Accepted Answer

qwen3-coder 30B-A3B MoE needs about 18.98GB working set (weights 18.6GB + KV cache 0.38GB at 4k context). The AMD Instinct MI355X 288GB offers ~288GB usable — verdict: full. 19.0 GB needed of 288 GB usable — headroom for context

Question 2

How fast is qwen3-coder 30B-A3B MoE on the AMD Instinct MI355X 288GB?

Accepted Answer

~1200–1900 tok/s (active-expert estimate)

computed roofline: 8000 GB/s × 0.28–0.45 efficiency window / 1.9GB (Q4_K_M) — bands, never points.

MoE: file is 18.6GB (that decides fit) but only ~3B of experts activate per token (~1.9GB streamed) — band below is the active-expert estimate, optimistic; community MoE rows run several× the naive bandwidth math

Question 3

What should I run qwen3-coder 30B-A3B MoE on instead?

Accepted Answer

Tesla V100 32GB (used) runs it fully ($640).

Can the AMD Instinct MI355X 288GB run qwen3-coder 30B-A3B MoE?

Run qwen3-coder 30B-A3B MoE on instead