Skip to content

triton-ascend-case-reduction-amax-small

极小规模归约(amax)优化:单核处理(grid=1)优于多核并行(2.16us vs 3.51us),避免并行化带来的调度开销,适用于数据规模很小(<1000元素)的归约场景

Repository Source folder

Details

Path
akg_agents/python/akg_agents/op/resources/skills/triton-ascend/cases/triton-ascend-case-reduction-amax-small/SKILL.md
Dependencies
3

FAQ