Analyze Triton kernels for Ascend NPU and generate optimized versions.
npx skills add https://github.com/Just-it/AscendOpGenAgent --skill latency-optimizer
Please help me install this Agent Skill. Skill: latency-optimizer Source: https://github.com/Just-it/AscendOpGenAgent/tree/main/skills/latency-optimizer Command: npx skills add https://github.com/Just-it/AscendOpGenAgent --skill latency-optimizer
Efficiently identify and apply performance optimizations to Triton kernels for Ascend NPU, ensuring correctness and preserved accuracy.
Provide the Triton kernel file path to latency-optimizer to receive an optimized kernel and validation notes.
High-intent search queries and answers about installing and using this skill.
To optimize Triton kernel performance for Ascend NPU, you provide the kernel file path for analysis. It identifies bottlenecks in patterns like constexpr parameters and int32 vectorization, then generates an optimized version with numerical equivalence notes.
Performance bottlenecks in Triton kernels on Ascend NPU often stem from constexpr parameters, int32 vectorization, load-order, and vector comparisons. The optimizer detects these specific patterns and generates targeted improvements.
Yes, validating accuracy after optimizing Triton ops for Ascend NPU is built into the process. The optimizer provides comparison reports and notes to ensure functional and numerical equivalence between original and optimized code.
The Ascend NPU Triton optimizer currently operates under a fixed configuration with PyTorch as the framework. It specifically processes Triton DSL code targeting the Ascend backend, ensuring optimized performance within this specific environment.
Your Triton kernel might run slow on Ascend NPU due to suboptimal load-order, unoptimized int32 vectorization, or inefficient constexpr parameters. Analyzing the code file identifies these specific bottlenecks and generates a faster version.
AscendGenerate Triton operator design documents for Ascend NPU.
AscendOptimize Triton operator performance on Ascend NPU by addressing UB overflow and tiling strategies.
xchang1121Apply optimization patterns to Triton Ascend kernels for improved performance.
mindspore-aiOptimize Triton CUDA kernels with performance strategies and API usage rules.
