PGI Community Edition

The PGI Community Edition with OpenACC offers scientists and researchers a quick path to accelerated computing with less programming effort. By inserting compiler “hints” or directives into your C, C++ or Fortran code, the PGI OpenACC compiler will offload and run your code on the GPU and CPU.

In addition to the PGI OpenACC compilers, the PGI Community Edition includes GPU-enabled libraries and developer tools to help you with your GPU acceleration effort.

Get Your Free PGI Community Edition Today

DOWNLOAD

  • Simple: Minimum code modifications to start on GPUs
  • Powerful: Up to 10x faster application performance
  • Portable: Same source code runs on CPUs and GPUs

Features

The PGI Community Edition includes a complete set of developer tools to help you achieve significant application acceleration with a minimum amount of coding effort. Featuring the PGI Accelerator Fortran/C/C++ compilers for Linux x86-64, Linux OpenPOWER, or macOS, it supports OpenACC 2.5 running on GPUs and OpenACC 2.5, CUDA Fortran and OpenMP 3.1 running on the CPU host.

Additional tools include:
  • PGPROF Performance Profiler to identify functions and loops that will run faster on GPUs.
  • GPU-enabled Libraries to use for already accelerated algorithms.
  • OpenACC Code Samples to get started with simple and real-world examples

Additional resources and documentation are available on the PGI website and at OpenACC.org.

You can also sign up for OpenACC hands-on labs and receive two free 90-minute, on-demand training sessions to quickly learn and master OpenACC techniques.

Application Acceleration

LS-DALTON: Benchmark on Oak Ridge Titan Supercomputer, AMD CPU vs Tesla K20X GPU. Test input: Alanine-3 on CCSD(T) module..
COSMO: additional information here.
NICAM: Benchmark on TiTech TSUBAME 2.5, Westmere CPU vs. K20X, additional information here.