Qwen3-Coder 30B (3B active)
on AMD Radeon 8060S integrated GPU (Strix Halo, gfx1151) · 1bit engine lean ROCm / ROCmFPX (W4A4)
Summary
User reports Qwen3-Coder 30B-A3B at 79.2 t/s generation and 2,156 t/s prompt processing on one Strix Halo integrated GPU.
Setup is ROCmFPX/lean ROCm (HIP) with rotated Q4_0 W4A4 attention and experts, GGML_W4A4_TENSORS=all, no MTP or drafter identified.
The figures are separate native llama-bench phases, -r 3, pp512 2,156 t/s and tg128 79.2 t/s, not a client request or TTFT; decode does not use W4A4.
Compare this run
Medians of plain runs: one device, one request, no speculative decoding, the whole model in the device's memory. The report above is one setup, so read it beside them and not as one of them.
Qwen3-Coder 30B · 3B active at 4-bit by hardware
- 107.0 t/s · one run
- 70.3 t/s · 2 runs
- 69.9 t/s · 6 runs
- 66.1 t/s · one run
Bars are medians on one scale, 0 to 107.0 tokens/s. Each dot is one run and opens its report. Press ▶ to watch text arrive at a row's speed.