inference/research
verifiedmedium · 0.75updated 2026-10-14

Tenstorrent Wormhole n300 Qwen2.5-7B decode (vendor-measured, b32 TP2)

22.1 tok/s per user at batch-32; 707 tok/s aggregate

Fields

id
benchmarks/comm-tt-wormhole-n300-qwen25-7b-decode
Metric
tok_s_per_user
Value
22.1
Unit
tok/s per user at batch-32; 707 tok/s aggregate
Model
Qwen2.5-7B
Methodology
Vendor tt-metal model matrix, batch-32 decode ISL=128, n300 TP=2
Measured by
vendor
Thermal state
unknown

Cross-references