skill
职业分类
描述
更新
benchmark-harness
计算机程序员
Use when creating or revising CUDA benchmark runners and result artifacts.
2026-03-17
cuda-env-audit
网络与计算机系统管理员
Use when CUDA availability, compiler setup, or profiler visibility may be broken or inconsistent.
2026-03-17
flashattention-kernel-design
软件开发工程师
Use when designing or reviewing handwritten CUDA FlashAttention kernels.
2026-03-17
gemm-kernel-design
软件开发工程师
Use when designing or reviewing handwritten CUDA GEMM kernels.
2026-03-17
ncu-profiling
软件开发工程师其他计算机职业
Use for kernel-level analysis with Nsight Compute after a benchmark is reproducible.
2026-03-17
nsys-timeline
其他计算机职业
Use for stream overlap, launch overhead, and end-to-end CUDA timeline analysis with Nsight Systems.
2026-03-17
roofline-analysis
计算机程序员
Use when deciding whether a CUDA kernel is likely memory-bound or compute-bound.
2026-03-17