Explore GPU parallelization in HPX's implementation of parallel STL in this CppNow conference talk. Dive into the integration of GPU computing power with C++ applications using executors and modern GPGPU frameworks. Learn about the challenges of GPU programming and how HPX addresses them through C++AMP and SYCL standards. Discover the approach of extending existing implementations with GPU executors, minimizing developer work on data transfers and synchronization. Examine the integration of HPX with Clang-based compilers supporting AMP and SYCL standards. Gain insights into the potential of combining mature C++ projects with GPU parallelization techniques for enhanced performance and flexibility in algorithm execution.
Overview
Syllabus
Marcin Copik: HPX and GPU parallelized STL
Taught by
CppNow