Trendoday

AscendGYM: A Cost-Aware Kernel Optimization Benchmark for Real-World Workloads on Ascend NPUs

Kernel optimization is essential for efficient large language model (LLM) training and inference on Ascend NPUs. Existing Ascend kernel benchmarks primarily focus on kernel generation, with less empha

  • Upvotes83
–Rank now
#2Peak rank
2Days on list
359Score · 7d
Nov 18, 2026First seen
Rank · 7 days

About this paper

AscendGYM provides a benchmark for assessing kernel optimization capabilities on Ascend NPUs using real-workload test cases, two evaluation modes, and a Speedup-Token Score that balances speedup against token cost.

Kernel optimization is essential for efficient large language model (LLM) training and inference on Ascend NPUs. Existing Ascend kernel benchmarks primarily focus on kernel generation, with less emphasis on the ability of agents to optimize existing implementations under input configurations drawn from real workloads. To address this requirement, we introduce AscendGYM, a benchmark whose initial suite comprises 48 kernels across three difficulty levels. Each kernel is paired with test cases whose data types and tensor shapes are collected from real-world training and inference tasks. Consequently, performance improvements on AscendGYM map directly to the corresponding workloads. In terms…

large language model (LLM)Ascend NPUskernel optimizationkernel benchmarkskernel generationagentsinput configurationstensor shapesmodel evaluationmodel-harness evaluation
Upvotes
83
Authors
27
Published
Nov 13, 2026

Summary and abstract excerpt via Hugging Face Papers ↗ · arXiv.

Trending history

AscendGYM: A Cost-Aware Kernel Optimization Benchmark for Real-World Workloads on Ascend NPUs was first seen on Hugging Face · Daily papers on Nov 18, 2026. Since then it has appeared in 12 checks on 2 different days. Its best position was #2 of 30, reached on Nov 18, 15:28 UTC. It dropped off the list after Nov 19, 05:28 UTC.

Rank history

Position on Hugging Face · Daily papers at every check (lower is better).

#1#11#20#30 07:2812:5818:2823:5805:28 Nov 18, 07:28 UTC · #3Nov 18, 09:28 UTC · #4Nov 18, 11:28 UTC · #4Nov 18, 13:28 UTC · #3Nov 18, 15:28 UTC · #2Nov 18, 17:28 UTC · #2Nov 18, 19:28 UTC · #2Nov 18, 21:28 UTC · #2Nov 18, 23:28 UTC · #2Nov 19, 01:28 UTC · #2Nov 19, 03:28 UTC · #2Nov 19, 05:28 UTC · #2

Last 14 days

Best rank per day (UTC).

Nov 9today

Metric history

Platform metric as reported by Hugging Face.

0306091 07:2812:5818:2823:5805:28 Nov 18, 07:28 UTC · 83Nov 18, 09:28 UTC · 86Nov 18, 11:28 UTC · 88Nov 18, 13:28 UTC · 90Nov 18, 15:28 UTC · 90Nov 18, 17:28 UTC · 91Nov 18, 19:28 UTC · 91Nov 18, 21:28 UTC · 91Nov 18, 23:28 UTC · 91Nov 19, 01:28 UTC · 91Nov 19, 03:28 UTC · 91Nov 19, 05:28 UTC · 91

Questions

Is AscendGYM: A Cost-Aware Kernel Optimization Benchmark for Real-World Workloads on Ascend NPUs trending on HF Daily Papers right now?

Not at the moment. It was last on the list on Nov 19, 05:28 UTC, after 2 days of trending.

What is the highest rank AscendGYM: A Cost-Aware Kernel Optimization Benchmark for Real-World Workloads on Ascend NPUs reached?

#2 on Hugging Face · Daily papers, on Nov 18, 15:28 UTC. Trendoday checks this list every 120 minutes.

How long has AscendGYM: A Cost-Aware Kernel Optimization Benchmark for Real-World Workloads on Ascend NPUs been trending?

It was first seen on Nov 18, 2026 and has been on the list on 2 days (12 checks) since.