NVIDIA Jetson page: Thor vLLM Qwen3 32B, concurrency 8, 79.1 tok/s aggregate
79.1 tok/s AGGREGATE across max concurrency 8
Fields
id
benchmarks/nv-thor-vllm-qwen3-32b-c8-aggregateMetric
tps_aggregate
Value
79.1
Unit
tok/s AGGREGATE across max concurrency 8
Model
Qwen3 32B
Methodology
NVIDIA Developer 'Jetson Benchmarks' page, Jetson AGX Thor Benchmarking Results table, read 2026-10-17. Jetson AGX Thor Developer Kit, JetPack 7.0, CUDA 13.0, TensorRT 10.13, vLLM (version not stated), input sequence length 2048, output sequence length 128. Weight dtype is not stated. The page gives no run count, no variance and no benchmark command, so confidence is below the MLPerf logs. Max concurrency 8 column, aggregate throughput.
Measured by
vendor
Reproducible
yes
notes
Vendor-measured. Weight quantization is not disclosed on the page, so format_id is left null rather than guessed. VLM rows from the same table were not filed.
Notes
Vendor-measured. Weight quantization is not disclosed on the page, so format_id is left null rather than guessed. VLM rows from the same table were not filed.