Parallel Programming

The parallel programming track connects LPPC with high-performance computing, systems, computer architecture, and experimental measurement. It works with correctness, speedup, concurrency, distribution, and responsible use of computational resources.

Marathon of Parallel Programming

The Marathon of Parallel Programming takes place in event contexts such as ERAD and SSCAD, formerly WSCAD, and works with parallel and distributed programming with a focus on correctness and performance gain. The contest often starts from sequential solutions and requires parallel versions evaluated by speedup.

• Target audience: undergraduate and graduate students interested in high-performance computing, systems, and architecture.
• Format: teams, problems with a sequential reference solution, and performance evaluation of the parallel or distributed version.
• Common technologies and languages: C, C++, OpenMP, MPI, Pthreads, and CUDA, depending on the edition.
• Level: advanced.
• Expected preparation: participants need C or C++, architecture basics, concurrency, memory, parallelism, and performance measurement.
• Links: current edition · 2024 edition · ERAD/RS 2024 · track resources

How To Contribute

Contributions may include problems with sequential solutions, reference parallel versions, measurement criteria, execution-script review, workshop preparation, and support in practical meetings. See Individuals and professionals.