cfg-0087
citable URL: https://halobench.com/records/cfg-0087/ — this address never moves; the anchor /records/#cfg-0087 keeps resolving
| model | NVIDIA-Nemotron-3-Super-120B-A12B-UD-Q4_K_M-00001-of-00003.gguf · UD-Q4_K_M |
| engine | ggml-org/llama.cpp 3653e6d · rocm · host aihydra (igpu) |
| flags | -ngl 999 -fa 1 -b 2048 -ub 512 -ctk f16 -ctv f16 -t 16 --load-mode none |
| template | not recorded at test time |
| tree | upstream — stock |
Capability basis
unmeasured — performance numbers on this config stand on a guard alone, not an established capability
Runs on this config (4)
| run | date | kind | suite | metrics |
|---|---|---|---|---|
| run-0308 | 2026-08-16 | performance | llama-bench@1 | prefill_tps=271.546 |
| run-0309 | 2026-08-16 | performance | llama-bench@1 | decode_tps=16.7065 |
| run-0310 | 2026-08-16 | performance | llama-bench@1 | prefill_tps=244.767 |
| run-0311 | 2026-08-16 | performance | llama-bench@1 | decode_tps=16.3962 |
Cited by — computed at build time, never stored
model pages nemotron3-super
docs lab-site-design