Question 1

Can the NVIDIA H200 141GB run qwen3-coder 480B A35B cloud-class?

Accepted Answer

qwen3-coder 480B A35B cloud-class needs about 290.97GB working set (weights 290GB + KV cache 0.97GB at 4k context). The NVIDIA H200 141GB offers ~141GB usable — verdict: no. 291.0 GB vs 141 GB usable — does not fit

Question 2

How fast is qwen3-coder 480B A35B cloud-class on the NVIDIA H200 141GB?

Accepted Answer

~7.6–13 tok/s on the NVIDIA H200 141GB. computed roofline: 4800 GB/s × 0.46–0.76 efficiency window / 290.0GB (Q4_K_M) — bands, never points.

Question 3

What should I run qwen3-coder 480B A35B cloud-class on instead?

Accepted Answer

ASUS ExpertCenter Pro ET900N G3 (GB300) runs it fully ($60000).

Can the NVIDIA H200 141GB run qwen3-coder 480B A35B cloud-class?

Run qwen3-coder 480B A35B cloud-class on instead