BIPP - Bluebild Imaging++

Image synthesis in radio astronomy is done with interferometry, a powerful technique allowing observation of the sky with antenna arrays with otherwise inaccessible angular resolutions and sensitivities. The Bluebild algorithm offers a novel approach to image synthesis, leveraging fPCA to decompose the sky image into distinct energy eigenimages. Bluebild Imaging++ is an HPC implementation of Bluebild.

Requirements

Bipp requires the following:

• C++17 compliant compiler
• CMake 3.11 and later
• BLAS and LAPACK library like OpenBLAS or Intel MKL
• fiNUFFT

Bipp can be configured with additional features (check the CMake options below). The optional requirements are:

• Python header files and pybind11 for building the Python interface
• CUDA 9.0 and later for Nvidia GPU hardware
• ROCm 5.0 and later for AMD GPU hardware
• cuFINUUFT with type 3 transform support with CUDA or ROCm enabled. Currently, only available in a fork (t3_d3 branch): https://github.com/AdhocMan/cufinufft/tree/t3_d3
• Umpire for advanced memory management
• MAGMA for an alternative eigensolver implementation on GPU. Currently mandatory with ROCm
• VC for improved CPU performance through customized vectorization

The Python module has the following dependencies:

• numpy
• astropy
• matplotlib
• tqdm
• pyproj
• scipy
• pandas
• healpy
• casacore

Installation

Bipp uses CMake to configure the build.

CMake options

Bipp can be configured with the following options:

Option	Values	Default	Description
BIPP_PYTHON	ON, OFF	ON	Build Python interface
BIPP_OMP	ON, OFF	ON	Enable multi-threading with OpenMP
BIPP_VC	ON, OFF	OFF	Use the VC library for vectorization
BIPP_GPU	OFF, CUDA, ROCM	OFF	Select GPU backend
BIPP_MAGMA	ON, OFF	OFF	Use MAGMA as eigensolver on GPU
BIPP_BUILD_TESTS	ON, OFF	OFF	Build test executables
BIPP_INSTALL	LIB, PYTHON, OFF	LIB	Set installation target
BIPP_UMPIRE	ON, OFF	OFF	Use the UMPIRE library for memory allocations
BIPP_BUNDLED_LIBS	ON, OFF	ON	Download and build spdlog, pybind11, googletest and json library.
BIPP_INSTALL_LIB	ON, OFF	ON	Add library to list of install targets.
BIPP_INSTALL_PYTHON	ON, OFF	ON	Add python module to list of install targets.
BIPP_INSTALL_LIB_SUFFIX	string	lib or lib64	Installation path suffix appended to CMAKE_INSTALL_PREFIX for library target
BIPP_INSTALL_PYTHON_PREFIX	string		If set, used instead of CMAKE_INSTALL_PREFIX for python module target.
BIPP_INSTALL_PYTHON_SUFFIX	string	platlib	Installation path suffix for python module target. If "platlib", the python platlib path will be used.

Some useful general CMake options are:

Option	Description
CMAKE_PREFIX_PATH	Semicolon separated list of search paths for external libraries
CMAKE_INSTALL_PREFIX	Path to installation target directory
BUILD_SHARED_LIBS	Build shared libraries when enabled (ON). Static libraries otherwise (OFF)
CMAKE_CUDA_ARCHITECTURES	Semicolon separated list of CUDA architectures to compile for
CMAKE_HIP_ARCHITECTURES	Semicolon separated list of HIP architectures to compile for

Manual Build

The build process follows the standard CMake workflow.

To install a minimal build of the library without Python support:

mkdir build
cd build
cmake .. -DBIPP_PYTHON=OFF -DCMAKE_INSTALL_PREFIX=/usr/local -DBIPP_INSTALL=LIB
make -j8 install

To build bipp with Python support and install the python module to custom directory:

mkdir build
cd build
cmake .. -DBIPP_PYTHON=ON -DBIPP_INSTALL=PYTHON -DCMAKE_INSTALL_PREFIX=${path_to_install_to} -DBIPP_PYBIND11_DOWNLOAD=ON
make -j8 install
export PYTHONPATH=${path_to_install_to}:$PYTHONPATH

Python - Pip

Bipp uses skbuild to build the Python module with CMake and Pip. The CMake options can be set through environment variables. Example:

BIPP_GPU=CUDA CMAKE_PREFIX_PATH="${path_to_finufft};${path_to_cufinufft};${CMAKE_PREFIX_PATH}" python3 -m pip install .