Code exmaples of the book "Parallel C++: Efficient and Scalable High-Performance Parallel Programming".
The folder notebooks contains a Jupyter notebooks for each chapter of the book. In each notebook are all code examples of the Chapter. To run the notebooks a Docker image stevenrbrandt/cxxxex-src is available. To run the Docker image on your local computer, please use the following commands
curl -LO https://raw.githubusercontent.com/ stevenrbrandt/CxxExplorer/master/docker-compose.yml
docker-compose up -dAfter the Docker image is started, a tab with the JupyterHub instance will be opened in your browser. For more details, we refer to Chapter 1 of the book. Recall to use the C++ 17 cling kernel to compile C++ code using the C++ Explorer [1].
The examples using the Message Passing Interface (MPI) or OpenMP are not working in the Jupyter Notebooks. In addition, the distrbuted HPX examples are not working within Jupyter Notebooks. Therefore, the code is available as C++ files in the cpp folder.
The slurm scripts to run the benchmarks are available in the scripts folder.
The code for the benchmarks which are not discussded in detail in the book are available in the benchmark folder.
The benchmarks can be build as
cd benchmark
mkdir build
cd build
cmake -DWITH_MPI=ON ..
make and the follwoing options are availabe
WITH_MPI- Builds the MPI examples (Default OFF): Requires MPIWITH_HPX- Builds the HPX examples (Default OFF): Requires HPX compiled with networkingWITH_HPX_SIMD- Builds the HPX example with SIMD (Default OFF): Requires HPX compiled with SIMDWITH_HPX_COROUTINES- Builds the HPX examples with COROUTINES (Default OFF): Requires HPX compiled with C++ 20
- Diehl, Patrick, and Steven R. Brandt. "Interactive C++ code development using C++ Explorer and GitHub Classroom for educational purposes." Concurrency and Computation: Practice and Experience 35.18 (2023): e6893. 10.1002/cpe.6893