kernel-microbenchmark Skill

Build, debug, and interpret vLLM GPU kernel microbenchmarks for CUDA, Triton, and CuteDSL, including CUPTI timing, correctness checks, generated-code inspection, multi-GPU measurements, and SOL sanity checks.

Install

npx skills add 'https://github.com/vllm-project/vllm/tree/main/.agents/skills/kernel-microbenchmark'

Download bundle ↓

agents/openai.yamlBrowse 4 files

49 tokens

233 bytes
Token encoding: o200k_base
Snapshot da3c07b

← Back to SKILL.md

1interface:2  display_name: "Kernel Microbenchmark"3  short_description: "Build and debug reliable GPU kernel microbenchmarks."4  default_prompt: "Use the kernel microbenchmark workflow to build or debug a vLLM GPU kernel benchmark."5

← All vLLM skills Project instructions →