Parallel Range Algorithms - The Evolution of Parallelism in C++

Explore the cutting-edge development of Parallel Range Algorithms in this comprehensive conference talk from C++Now 2025, delivered by Ruslan Arutyunyan, one of the authors of the active C++ standard proposal P3179 that aims to introduce these algorithms to C++26. Discover how Parallel Range Algorithms combine the power of C++ Ranges with parallel execution to improve code clarity, expressiveness, and performance in modern C++ development. Learn about the key differences between these new algorithms and existing C++17 parallel algorithms and serial range algorithms, while examining critical design choices including bounded ranges, range-as-the-output approaches, and various corner cases. Understand the performance and safety implications of implementing Parallel Range Algorithms in your codebase, and explore their applications in heterogeneous computing environments based on oneAPI DPC++ library (oneDPL) experience. Gain insights into the technical design decisions behind this proposal through practical code examples and discussions on how these algorithms enable more expressive and scalable parallelism. Benefit from the expertise of a Senior Middleware Development Engineer who specializes in parallel and threading runtimes, leads the development of oneAPI DPC++ library (oneDPL), contributes to oneAPI Threading Building Blocks (oneTBB), and serves as co-chair of SG1 (Concurrency and Parallelism) in the C++ Standard Committee, making this session accessible to both library authors and everyday C++ developers interested in the future of parallel programming.