Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Inference Throughput on Qwen3 Query Projection Module NVIDIA A40

80.63Throughput (k tokens/sec)

TTQ (r = 0)

3.909223.827143.74563.6629Mar 11, 2026
Updated 26d ago

Evaluation Results

MethodLinks
2026.03
80.63
2026.03
78.68
2026.03
75.76
2026.03
73.98
2026.03
73.7
2026.03
73.32
2026.03
71.6
2026.03
69.11
2026.03
66.67
2026.03
59.08
2026.03
57.58
2026.03
54.61
2026.03
49.67
2026.03
41.29
2026.03
38.91
2026.03
37.52
2026.03
36.52
2026.03
32.68
2026.03
30.32
2026.03
30.32
2026.03
25.28
2026.03
24.94
2026.03
23.78
2026.03
21.38
2026.03
19.34
2026.03
18.28
2026.03
17.43
2026.03
14.1
2026.03
10.43
2026.03
6.86