CANN Bench

What is CANN Bench?

Proposes a benchmark for AI-generated operator code on Huawei's Ascend NPU, covering 53 operators and 1060 test cases. Evaluation uses a three-dimensional…

Released

2026-07-08

Evaluates

Agents, Language & Knowledge, Software & AI Compute

Openness

unknown

Importer

Claire Radar

Review status

unreviewed

Reported scores

0

Source provenance

• Original evidence https://arxiv.org/abs/2607.20518

CANN Bench paper

• Paper https://arxiv.org/abs/2607.20518

Which sources cite CANN Bench?

• CANN Bench

Related Agents, Language & Knowledge, Software & AI Compute benchmarks

• $\tau^\tau$-Bench: An Environment for End-To-End, Realistic Agent Construction
• 5-Dialects-BN
• A Large-Scale Dataset and Benchmark
• A3S-Bench
• AARRI-Bench
• ABC-Bench
• Active-SWE
• ActiveSciBench