Decode · tok/s
177.6
Prefill
—
tok/100W
77.2
Qwen3.8-27B-NVFP4 na A5000 z kontekstem 262K … Szybkość generowania: 177.57 tokens/s
Rank is the quoted TensorFold 177.57. Prefill “ponad 7700” skipped (over, not a point).
RTX A5000 · NVFP4 · TensorFold
Qwen3.8-27B-NVFP4 na A5000 z kontekstem 262K … Szybkość generowania: 177.57 tokens/s
Rank is the quoted TensorFold 177.57. Prefill “ponad 7700” skipped (over, not a point).