-
Notifications
You must be signed in to change notification settings - Fork 0
Building
Hugo edited this page Jan 8, 2026
·
1 revision
- CMake (3.10 or higher)
- C Compiler (GCC, Clang, or MSVC)
- (Optional) CUDA Toolkit (for GPU support)
-
Create a build directory:
mkdir build && cd build
-
Configure the project:
cmake ..
-
Build the project (Release mode recommended):
cmake --build . --config Release
You can configure the build by passing -D<OPTION>=<VALUE> to CMake.
| Option | Default | Description |
|---|---|---|
NOCTA_BUILD_EXAMPLES |
ON | Build example programs |
NOCTA_ENABLE_SIMD |
ON | Enable AVX2 SIMD optimizations |
NOCTA_ENABLE_OPENMP |
ON | Enable OpenMP multi-threading |
NOCTA_ENABLE_CUDA |
OFF | Enable CUDA/cuBLAS GPU acceleration |
To enable GPU acceleration, ensure you have the CUDA Toolkit installed.
mkdir build && cd build
cmake .. -DNOCTA_ENABLE_CUDA=ON \
-DCUDAToolkit_ROOT=/usr/local/cuda-12.x/ \
-DCMAKE_CUDA_COMPILER=/usr/local/cuda-12.x/bin/nvcc
cmake --build . --config Release