Home › Evidence › Records › cfg-0051

cfg-0051

backfilled
citable URL: https://halobench.com/records/cfg-0051/ — this address never moves; the anchor /records/#cfg-0051 keeps resolving
modelQwen3.6-35B-A3B-UD-Q4_K_XL.gguf · UD-Q4_K_XL
engineggml-org/llama.cpp ed89854 · rocm · host aihydra (igpu)
flags-ngl 999 -fa 1 -b 2048 -ub 512 -ctk f16 -ctv f16 -t 16
templatenot recorded at test time
treeupstream — stock
aged evidence — reconstructed from the archive. Ingested from llama-bench, which measures throughput and not footprint: the memory block, chat-template hash and model revision are absent by construction. Peak memory comes from a Tier 0 capability probe, not from this tool. Everything present here — build commit, backend, KV quant, batch sizes, flash attention — is machine-reported.

Capability basis

unmeasured — performance numbers on this config stand on a guard alone, not an established capability

Runs on this config (6)

rundatekindsuitemetrics
run-02062026-08-14performancellama-bench@50prefill_tps=1075.40898
run-02072026-08-14performancellama-bench@50decode_tps=54.001268
run-02082026-08-14performancellama-bench@50prefill_tps=257.931381
run-02092026-08-14performancellama-bench@50decode_tps=29.844321
run-02102026-08-14performancellama-bench@50prefill_tps=607.281338
run-02112026-08-14performancellama-bench@50decode_tps=44.957416