mirror of
https://github.com/NVIDIA/nvbench.git
synced 2026-04-01 14:19:00 +00:00
807917cf54
Add sphinx-combined folder that builds combined C++ & Python docs Fixed relative text alignment in docstrings to fix autodoc warnigns Renamed cuda.bench.test_cpp_exception and cuda.bench.test_py_exception functions to start with underscore, signaling that these functions are internal and should not be documented Account for test_cpp_exceptions -> _test_cpp_exception, same for *_py_* Fix cpp_benchmarks, add py_benchmarks 1. Fixed xrefs in docs/sphinx-combined/cpp_benchmarks.md, which is built on top of docs/benchmarks.md Added level-1 heading, and pushed existing headings one level down. 2. Added py_benchmarks.md to document benchmarking of Python scripts. 3. Rearranged entries in index.rst so that overview documents come before API enumeration. Make sure to reset __module__ of reexported symbols to be cuda.bench Enumerate free functions in nvbench:: namespace Tweak to index.rst intro sentence and title Changed title, fixed references, added intro borrowed from README Fix punctuation in one of the itemlist item text Hide TOC from the index page. It is too long and confusing
CUDA Kernel Benchmarking Package
This package provides a Python API to the CUDA Kernel Benchmarking
Library NVBench.
Installation
Install from PyPi
pip install cuda-bench[cu13] # For CUDA 13.x
pip install cuda-bench[cu12] # For CUDA 12.x
Building from source
Ensure recent version of CMake
Since nvbench requires a rather new version of CMake (>=3.30.4), either build CMake from sources, or create a conda environment with a recent version of CMake, using
conda create -n build_env --yes cmake ninja
conda activate build_env
Ensure CUDA compiler
Since building NVBench library requires CUDA compiler, ensure that appropriate environment variables
are set. For example, assuming CUDA toolkit is installed system-wide, and assuming Ampere GPU architecture:
export CUDACXX=/usr/local/cuda/bin/nvcc
export CUDAARCHS=86
Build Python project
Now switch to python folder, configure and install NVBench library, and install the package in editable mode:
cd nvbench/python
pip install -e .
Verify that package works
python test/run_1.py
Run examples
# Example benchmarking numba.cuda kernel
python examples/throughput.py
# Example benchmarking kernels authored using cuda.core
python examples/axes.py
# Example benchmarking algorithms from cuda.cccl.parallel
python examples/cccl_parallel_segmented_reduce.py
# Example benchmarking CuPy function
python examples/cupy_extract.py