Simplify Coding Across Architectures with DPC++

Get an overview of Data Parallel C++ (DPC++), an open, standards-based, cross-architecture language that simplifies heterogeneous programming across CPUs, GPU, FPGAs, and specialized accelerators without proprietary lock-in, providing developers freedom of choice in hardware.

DPC++ highlights:

  • Provides developers freedom of choice in hardware
  • Is an open-community project based on C++ and contains SYCL to support parallelism
  • Interoperates with other languages (C, C++, Fortran, Python, and others) and parallel models (OpenMP, MPI, and others)
  • Allows code reuse across hardware targets while permitting custom tuning for a specific accelerator
  • Delivers native high-level language performance

Watch.

Get the software

  • DPC++ is included as part of the Intel® oneAPI Base Toolkit, a core set of tools and libraries for developing high-performance, data-centric applications across diverse architectures.
  • Try it in the Intel® DevCloud, where you can learn DPC++ basics and then use it to develop, test, and run workloads on a cluster of the latest Intel® hardware and oneAPI software.
Performance varies by use, configuration, and other factors. Learn more at www.Intel.com/PerformanceIndex.