Wormhole n150 tt-metal: Whisper distil-large-v3 batch-1 tok/s per user
105.0 tok/s per user at batch 1 (equals aggregate at batch 1)
Fields
id
benchmarks/ttmetal-n150-whisper-distil-large-v3-b1-tpsuMetric
tok_s_per_user
Value
105
Unit
tok/s per user at batch 1 (equals aggregate at batch 1)
Model
Whisper distil-large-v3
Methodology
VENDOR demo in tenstorrent/tt-metal models/README.md speech-to-text table (matrix last updated 2025-12-08). Whisper distil-large-v3, single n150 (Wormhole), batch 1, TT-Metalium v0.65.0-dev20251208. t/s/u 105.0 = 1/inter-token-latency, and at batch 1 the aggregate t/s is also 105.0. Per-request, not a concurrent aggregate.
Measured by
vendor
Reproducible
yes
notes
Vendor demo numbers, not a vLLM serving trace. The README states t/s = t/s/u * batch and that vLLM-server performance may differ. LLM rows use input sequence length 128 except Mamba, which accepts any length. Starred Blackhole figures carry the vendor's own caveat that the software is still under active optimization.
Notes
Vendor demo numbers, not a vLLM serving trace. The README states t/s = t/s/u * batch and that vLLM-server performance may differ. LLM rows use input sequence length 128 except Mamba, which accepts any length. Starred Blackhole figures carry the vendor's own caveat that the software is still under active optimization.