Developer Guide


Optimize Your Design

This chapter describes features and provides guidance on leveraging the functionalities of Data Parallel C++ (DPC++) to optimize your designs.
In general, the methods you use to improve the performance of your DPC++ kernels should achieve the following results:
  • Increase the number of parallel operations.
  • Increase the memory bandwidth of the implementation.
  • Increase the number of operations per clock cycle that kernels can perform in hardware.

Product and Performance Information


Performance varies by use, configuration and other factors. Learn more at