Data Parallel C++ (DPC++)

DPC++ = ISO C++ and Khronos SYCL* and community extensions

Learn the Basics of DPC++

Data Parallel C++ (DPC++) is an open alternative to single-architecture proprietary languages. It allows developers to reuse code across hardware targets (CPUs and accelerators such as GPUs and FPGAs) and also perform custom tuning for a specific accelerator. These code walkthroughs will introduce you to basic principles and practices of DPC++ programming. To continue learning about DPC++ and Intel’s integration of this new programming language into the oneAPI Toolkits, see More Ways to Learn.

1. Concepts and Functionality

The Vector Add sample demonstrates oneAPI concepts and functionally using the DPC++ programming language. You will learn about:

Device selectors targeting different accelerators including GPU and FPGA
Buffers and accessors
Queues
Data parallel kernel “parallel_for”

Go to Code Walkthrough

2. Unified Shared Memory (USM)

The Mandelbrot Set is a program that demonstrates oneAPI concepts and functionally using the DPC++ programming language. You will learn about:

Unified shared memory
Managing and accessing memory
Parallel implementation

Go to Code Walkthrough

More Ways to Learn

Try in the Intel® DevCloud for Free

Develop, run, and optimize your code in this cloud-based development sandbox with 120 days of full access. Access samples or run your own workloads.

Access oneAPI Software

Toolkits: Intel® oneAPI Base Toolkit and add-on toolkits for HPC, AI, IoT, and more
Tools and Libraries: Intel® oneAPI DPC++/C++ Compiler, Intel® oneAPI Math Kernel Library, Intel Distribution for Python*, Intel® VTune™ Profiler, Intel® Advisor, and more

Use the Latest Intel® Hardware