Strix Halo 8060S (HIP rocWMMA (ROCm)) Mistral Small 3.1 24B UD-Q4_K_XL: tg128 decode tok/s
13.8 tok/s (tg128, per-request batch-1)
Fields
id
benchmarks/strix-halo-8060s-mistral-small-31-udq4-hip-tg128Metric
decode_tok_s
Value
13.8
Unit
tok/s (tg128, per-request batch-1)
Model
Mistral Small 3.1 24B
Methodology
THIRD-PARTY MEASURED (lhl/strix-halo-testing llm-bench summary table). Hardware: Framework Desktop, AMD Ryzen AI Max+ 395 (Strix Halo) + Radeon 8060S iGPU, 128GB LPDDR5x-8000 unified memory. llama.cpp ~b5863, HIP rocWMMA (ROCm) backend, flags fa=1 hipBLASLt; llama-bench -fa 1 --mmap 0 -ngl 99, pp512 + tg128, per-request batch-1. Peak resident 14638 MiB. Numbers verbatim from the best-backend summary table.
Measured by
third_party
Reproducible
yes