PyOMP documentation

PyOMP implements OpenMP parallel CPU/GPU programming abstractions in Python using Numba.

Check out the Installation section for how to install, the Usage section on information regarding using PyOMP, and OpenMP support for the list of supported OpenMP abstractions.

Contents

• Installation
  • Compatibility
  • Additional options
• Usage
  • Syntax
    • Overview
    • OpenMP directives
    • OpenMP runtime functions
  • Examples
    • CPU parallelism example
    • GPU offloading example
• OpenMP support
  • OpenMP directives and clauses
    • barrier
    • critical
    • for
    • parallel
    • parallel for
    • single
    • task
    • taskwait
    • target
    • teams
    • distribute
    • teams distribute
    • target teams
    • target data
    • target enter data
    • target exit data
    • target update
    • target teams distribute
    • distribute parallel for
    • target teams distribute parallel for
  • OpenMP runtime functions
    • Thread and team information
    • Timing
    • Nested and hierarchical parallelism
    • Advanced features
    • Device and target offloading
  • Supported features and platforms
    • OpenMP and GPU offloading support
    • Device selection and querying
      • Discovering Available Devices
      • Finding devices by criteria
      • Querying device properties
      • Using device ids in target regions
    • Version and platform support
      • OpenMP parallelism support by platform
      • Platform details
      • Notes