lhl 8060S Qwen3-32B-Q8_0 Vulkan tg1 decode
6.43 tok/s
Fields
id
benchmarks/lhl-8060s-qwen3-32b-q8-0-vulkan-tg1Metric
decode_tok_s
Value
6.43
Unit
tok/s
Model
Qwen3-32B-Q8_0
Methodology
lhl strix-halo-testing llm-bench (github.com/lhl/strix-halo-testing, Qwen3-32B-Q8_0/results.jsonl). llama.cpp llama.cpp-vulkan, backend Vulkan, no -fa, ngl=99, tg1 batch-1 decode at depth 0 on Ryzen AI Max+ 395 (Radeon 8060S gfx1151). System RAM peak 37388MiB, avg power 45W, max temp 59.0C. Per-request batch-1.
Measured by
third_party
Reproducible
yes