CMake Options for NEST GPU
NEST GPU is installed with cmake
(at least version 3.17). In the simplest case, the commands:
cmake -DCMAKE_INSTALL_PREFIX:PATH=<nestgpu_install_dir> <nestgpu_source_dir>
make
make install
should build and install NEST GPU to nestgpu_install_dir
, which should be an absolute
path.
Options for configuring NEST GPU
NEST GPU allows for several configuration options for custom builds:
Set GPU architecture
See here to choose the best NVIDIA GPU architecture for your GPU card.
|
Specify the GPU compute architecture. [default=70] |
Change parallelization scheme
|
Build with MPI parallelization. [default=ON] |
|
Enable OpenMP multi-threading. Optionally set OMP compiler flag. [default=ON] |
Warning
Currently NEST GPU must be installed with MPI to avoid compilation errors.
External libraries
|
Build with ltdl library. To set a specific ltdl, give the install path. NEST GPU uses ltdl for dynamic loading of external user modules. [default=OFF] |
Compiler options
Explicit compiler flags can be given.
|
Enable user defined optimizations. When OFF, no ‘-O’ flag is passed to the compiler. Separate multiple flags by ‘;’. [default=ON (uses ‘-O3’)] |
|
Enable user defined warnings. Separate multiple flags by ‘;’. [default=ON (uses ‘-Wall’)] |
|
Enable user defined debug flags. Separate multiple flags by ‘;’. [default=OFF, when ON, defaults to ‘-g’] |
CUDA options
|
C++ standard to use for compilation. [default=OFF] |
|
Link additional libraries. Give full path. Separate multiple libraries by ‘;’. [default=OFF] |
|
Add additional include paths. Give full path without ‘-I’. Separate multiple include paths by ‘;’. [default=OFF] |
|
Additional defines, e.g. ‘-DXYZ=1’. Separate multiple defines by ‘;’. [default=OFF] |
|
Set a maximum amount of register used when compiling. [default=55] |
|
Options for ptxas compiling. Separate multiple flags by ‘;’. [default=OFF] |
Generic build configuration
|
Set a user defined version suffix. [default=’’] |