القائمة
Run matmul performance benchmarks. Use when user wants to measure TFLOPS, compare kernel performance, or verify correctness after code changes.
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
SM120 (Blackwell) CUDA expert. Use for wgmma/mma PTX inline assembly, TMA, narrow precision (FP4/FP6/FP8), and block-scaled GEMM development.
View and analyze build logs. Use when user wants to see build errors, check previous build output, or debug build failures.
Run LLM inference tests with Qwen or other models. Use when testing model loading, inference, CUDA Graph, or generation quality.
Run all checks including lint, typecheck, and tests. Use before creating PRs or for comprehensive validation.
CUDA kernel development workflow. Use when writing, testing, or optimizing GPU kernels. Follows the Edit-Build-Validate-Benchmark-Commit cycle.
Run Ruff linter and formatter on Python code. Use before commits or when checking code style and quality issues.
استنادا إلى تصنيف SOC المهني
| name | benchmark |
| description | Run matmul performance benchmarks. Use when user wants to measure TFLOPS, compare kernel performance, or verify correctness after code changes. |
Run comprehensive matmul benchmarks for all supported dtypes.
# Full benchmark
python scripts/benchmark.py
# Quick mode (fewer iterations)
python scripts/benchmark.py --quick
# Specific sizes
python scripts/benchmark.py --sizes 4096,8192
# TF32 kernel version
python scripts/benchmark.py --tf32-version v2
--sizes: Comma-separated matrix sizes (default: 2048,4096,8192)--quick: Fewer warmup/iterations for faster results--dtypes: Which dtypes to test (default: fp32,tf32,fp16,bf16)--tf32-version: v1 (WMMA) or v2 (PTX mma.sync, default)python scripts/benchmark.py [options]| Dtype | Target TFLOPS |
|---|---|
| FP32 | ~18 |
| TF32 | ~27 |
| FP16 | ~15 |
| BF16 | ~15 |
PYGPUKIT_ALLOW_TF32=1: Enable TF32 TensorCorePYGPUKIT_TF32_V2=1: Use PTX mma.sync kernel