lhl 8060S Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL Vulkan tg1 decode
19.9 tok/s
Fields
id
benchmarks/lhl-8060s-llama-4-scout-17b-16e-instruct-ud-q4-k-x-vulkan-tg1Metric
decode_tok_s
Value
19.9
Unit
tok/s
Model
Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL
Methodology
lhl strix-halo-testing llm-bench (github.com/lhl/strix-halo-testing, Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL/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 63314MiB, avg power 44W, max temp 73.0C. Per-request batch-1.
Measured by
third_party
Reproducible
yes