Strix Halo 8060S (HIP rocWMMA (ROCm)) Shisa V2 70B i1-Q4_K_M: pp512 prefill tok/s
94.7 tok/s (pp512, per-request batch-1)
Fields
id
benchmarks/strix-halo-8060s-shisa-v2-70b-q4km-hip-pp512Metric
prefill_tok_s
Value
94.7
Unit
tok/s (pp512, per-request batch-1)
Model
Shisa V2 70B
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 hipBLASLt; llama-bench -fa 1 --mmap 0 -ngl 99, pp512 + tg128, per-request batch-1. Peak resident 41522 MiB. Numbers verbatim from the best-backend summary table.
Measured by
third_party
Reproducible
yes