LeaseWeb: GPT-OSS-20B FP16 on NVIDIA L4, 58.6 tok/s decode
58.6 tok/s (batch-1 decode)
Fields
id
benchmarks/leaseweb-l4-gptoss20b-fp16-decodeMetric
decode_tok_s
Value
58.6
Unit
tok/s (batch-1 decode)
Model
gpt-oss-20b FP16
Methodology
THIRD-PARTY MEASURED (LeaseWeb blog, same article/rig as the EPYC-9334 rows). NVIDIA L4 GPU, FP16 model, 512-token prompt + 128 generated. API-level OpenLLM/llmperf measurement.
Measured by
third_party
Reproducible
no
notes
Same-rig comparison row; L4 vs EPYC-9334-dual ratio ~2.2x on this MoE model.
Notes
Same-rig comparison row; L4 vs EPYC-9334-dual ratio ~2.2x on this MoE model.