lhl 8060S GLM-4.5-Air-UD-Q4_K_XL Vulkan tg1 decode
22.89 tok/s
Fields
id
benchmarks/lhl-8060s-glm-45-air-ud-q4-k-xl-vulkan-tg1Metric
decode_tok_s
Value
22.89
Unit
tok/s
Model
GLM-4.5-Air-UD-Q4_K_XL
Methodology
lhl strix-halo-testing llm-bench (github.com/lhl/strix-halo-testing, GLM-4.5-Air-UD-Q4_K_XL/results.jsonl). llama.cpp llama.cpp-vulkan_amdvlk, backend Vulkan, no -fa, ngl=99, tg1 batch-1 decode at depth 0 on Ryzen AI Max+ 395 (Radeon 8060S gfx1151). System RAM peak 74356MiB, avg power 30W, max temp 56.0C. Per-request batch-1.
Measured by
third_party
Reproducible
yes