OpenMP Benchmark Comparison
SPECompG_base2012 relative performance as measured by PGI during the weeks of Oct. 9 and Nov. 3, 2015. SPEComp® is a registered trademark of the Standard Performance Evaluation Corporation (SPEC).  System:2 x Intel(R) Xeon(R) CPU E5-2698 v3 @ 2.30GHz (36 cores total) 128GB memory

On Intel Haswell CPUs with OpenMP, PGI delivers multicore performance more than 50% faster than the latest GCC compilers. That’s like buying a cluster with 50% more compute nodes. PGI compilers deliver world-class multicore CPU performance and accelerator programming features that can dramatically increase the performance of applications on GPU accelerators.

Run same code stream on CPU and GPU

Is your application 10s or 100s of thousands of lines of Fortran, C and C++ code? With OpenACC directives, you don’t have to parallelize all of it at once. You can identify hot loops and code regions using the PGI profiler, then incrementally parallelize and tune them one by one. OpenACC code remains 100% standard-compliant and portable to other compilers and platforms, and enables parallel processing on CPUs and GPUs using identical source code.

CloverLeaf Performance Comparison
We were extremely impressed that we can run OpenACC on a CPU with no code change and get equivalent performance to our OpenMP implementation.

The same code delivers up to 4X faster performance on a GPU. From a performance portability and code future proofing perspective, this is an excellent result.
Wayne Gaudin and Oliver Perks AWE, UK

CloverLeaf, a Lagrangian-Eulerian explicit hydrodynamics mini-application, is a small (4,500 line) lightweight application that is representative of the code used at the United Kingdom’s Atomic Weapons Establishment (AWE). Using OpenACC, performance on an NVIDIA K80 GPU is over 4X faster than a dual-socket 20-core Intel Ivy Bridge CPU, running the fully optimized code using a 960x960 mesh. The optimizations to the source code made during porting to the GPU using OpenACC improved the performance of the CPU code by more than 50%.

PGI Supports All Major HPC Platforms

HPC servers are quickly expanding beyond multicore x86 CPUs to OpenPOWER, ARM and GPU accelerators. PGI Fortran, C and C++ compilers and OpenACC are designed to deliver high performance on all of these processors. PGI compilers for x86, OpenPOWER and GPUs are available now, including OpenACC parallelization across all cores of a multicore CPU or a GPU. PGI and OpenACC deliver the performance you need today, and the flexibility you need tomorrow. PGI compilers can take you there.

PGI Profiler GUI

The PGI Profiler (a/k/a PGPROF®> is a powerful and easy-to-use interactive performance profiler for parallel programs written with OpenMP or OpenACC directives, or using CUDA. Use it to visualize and analyze the performance of your Fortran, C and C++ programs. The PGI Profiler can correlate execution time with procedures, source code and instructions, allowing you to quickly see where and how execution time is spent. Through resource utilization data and compiler feedback information, the PGI Profiler provides features that will help you understand why parts of your program have high execution times and how you can modify your source code or compiler options to improve performance. The PGI Profiler is included with all PGI products.

PGI Debugger GUI

The PGI graphical debugger for Fortran, C and C++ supports of debugging serial and parallel programs including MPI, OpenMP and hybrid MPI/OpenMP applications. The PGI Debugger can debug programs on SMP workstations, servers, distributed-memory clusters and hybrid clusters where each node contains multiple multicore x86 processors. It allows you to control threads or processes individually or in groups, and allows you to examine state down to the register level. The PGI Debugger is also included with all PGI products.

Buddy Bland, Project Director, Oak Ridge National Laboratory
Porting HPC applications from one platform to another is one of the most significant costs in the adoption of breakthrough hardware technologies. OpenACC for multicore x86 CPUs provides continuity and code portability from existing CPU-only and GPU- enabled applications from machines like Titan to all of DOE’s upcoming major systems as well as portability among those systems.
Buddy BlandProject Director at Oak Ridge National Laboratory

PGI Fortran is a native parallelizing/optimizing Fortran 2003 compiler with support for OpenMP 3.1 directive-based parallel programming for multi-core CPUs, and OpenACC 2.5 directive-based parallel programming for CPUs and accelerators. The PGI Fortran compiler includes CUDA Fortran extensions targeting NVIDIA CUDA-enabled GPUs.

PGI C/C++ are native parallelizing/optimizing ANSI C11 and GNU-compatible C++14 compilers. Both compilers implement full OpenMP 3.1 pragma-based parallel programming for multi-core CPUs, and OpenACC 2.5 pragma-based parallel programming for CPUs and accelerators. The PGI C++ compiler is substantially compatible with GNU GCC 4.8 g++ and includes full support for NVIDIA CUDA C extensions targeting multi-core CPUs as a compute device (CUDA-x86). PGI Fortran supports the industry standard ISO_C_BINDING, which allows for easy argument passing and procedure invocation between Fortran, C, and C++.

Note: Effective with the PGI 2016 release, the PGI C++ compiler is no longer supported on Microsoft Windows.

All PGI products include the OpenMP and MPI enabled parallel PGI Debugger that can debug up to 16 local MPI processes. PGI products also include the OpenMP and OpenACC parallel PGI Performance Profiler. All PGI products also include a precompiled MPI message passing library and precompiled OpenBLAS library.

PGI products for both macOS and Windows consists of command-level versions of the PGI compilers and both command-level and graphical versions of the PGI Debugger and PGI Performance profiler. An integrated development environment (IDE) is neither provided nor supported. PGI Visual Fortran fully integrates PGI Fortran compilers and tools into Microsoft Windows using Microsoft Visual Studio. PGI Visual Fortran is included with all PGI products for Windows.

PGI products target 64-bit x86 workstations, servers and clusters with one or more single core or multi-core microprocessors running Linux, macOS or Windows. In addition, PGI compiliers & tools tartget 64-bit OpenPOWER systems running Linux.

Technical Features

A partial list of technical features supported includes the following:

Features marked with an asterisk (*) are not currectly supported on OpenPOWER.

  • PGFORTRAN™ native OpenMP, OpenACC and auto-parallel Fortran 2003 compiler with CUDA extensions
  • PGCC® OpenMP, OpenACC and auto-parallel ANSI and K&R C11 compiler
  • PGC++® OpenMP, OpenACC and auto-parallel GNU 4.8 g++ compatible C++14 compiler with CUDA-x86 extensions (not available on Windows)
  • OpenMP and MPI parallel graphical debugger*
  • OpenMP and OpenACC parallel graphical performance profiler
  • Full 64-bit support on multi-core OpenPOWER and x86
  • Full support for OpenMP 3.1 on up to 256 cores
  • Preliminary support for OpenMP 4.5 for OpenPOWER CPUs (Fortran only)
  • Comprehensive OpenACC 2.5 support
  • PGI Unified Binary™ technology combines into a single executable or object file code optimized for multiple 64-bit x86 processors, and NVIDIA GPUss
  • Complete uniform development environment across 64-bit x86 processor-based systems running Linux, macOS or Windows and OpenPOWER processor-based systems running Linux
  • Comprehensive set of compiler optimizations including one pass interprocedural analysis (IPA)*, interprocedural optimization of libraries*, profile feedback optimization*, dependence analysis and global optimization, function inlining including library functions, vectorization, invariant conditional removal, loop interchange, loop splitting, loop unrolling, loop fusion and more.
  • Support for 64-bit integers (-r8/-i8 compilation flags)
  • Memory hierarchy and memory allocation optimizations including huge pages support
  • Auto-parallelization of loops specifically optimized for multi-core processors
  • Concurrent subroutine call support
  • Highly tuned Intel MMX and SSE intrinsics library routines (C/C++ only)
  • Tuning for non-uniform memory access (NUMA) architectures
  • Process/CPU affinity support in SMP/OpenMP applications
  • Support for creating shared objects on Linux, dynamic libraries on macOS and DLLs on Windows
  • Integrated cpp pre-processing
  • Cray/DEC/IBM extensions (including Cray POINTERs & DEC STRUCTURES/UNIONS); support for SGI-compatible DOACROSS in Fortran
  • Full support for Common Compiler Feedback Format compiler optimization listings
  • User modules support simplifies switching between multiple compiler environments/versions
  • C/C++ plug-in for Eclipse
  • Bundled precompiled libraries including ScaLAPACK (Linux & macOS), Open MPI (Linux only), MPICH (macOS) and MS-MPI library (Windows only)
  • Includes optimized 64-bit OpenBLAS (LAPACK/BLAS) math libraries
  • Supports multi-threaded execution with Intel Math Kernel Libraries (MKL) 10.1 and later on x86 processors
  • UNIX-compatible build/edit environment for Windows, including the BASH shell, vi editor, make, tar, gzip, sed, grep, awk, and over 100 other shell commands!
  • Interoperable with TotalView* (Linux only) and Allinea DDT*.
  • Interoperable with gcc, g77, g++ and gdb
  • Unconditional 30 day money back guarantee

System Requirements

  • Host Processor: 64-bit OpenPOWER, 64-bit x86 (including AMD64 and Intel 64) processor-based workstation or server with one or more single core or multi-core microprocessors.
    Accelerator (optional): NVIDIA CUDA-enabled GPU with compute capability 2.0 or later.
  • Operating System:
    • OpenPOWER Linux: Ubuntu 14.04, 14.10 and Red Hat Enterprise Linux 7.3.
    • x86-64 Linux: CentOS 5 or newer, OpenSuSE 11 or newer including OpenSuSE Leap 42.2, SUSE Linux Enterprise Server (SLES) 11 or newer, Red Hat Enterprise Linux 5 or newer, Fedora Core 6 or newer, or Ubuntu 12.04 or newer. Fully interoperable with versions of Linux using kernel revision 2.6 and glibc 2.5 or newer.
    • 64-bit Apple macOS version 10.9 Mavericks or newer and Xcode 5.1 or newer.
    • 64-bit Microsoft Windows 10/8.1/8/7 and Server 2016/2012/2008 R2.

    Please Note: 32-bit development was deprecated with the PGI 2016 release and is no longer available with the PGI 2017 release.

  • Memory: 16 MB or more.
  • Hard Disk: 1.5 GB during installation, 700 MB to hold installed software.
  • Peripherals: Mouse or compatible pointing device for use of optional graphical user interfaces.
  • Other: Adobe Acrobat Reader for viewing documentation.

Which releases of PGI compilers work with each Linux distribution?

Each release of Linux changes many things, primarily the glibc libraries, the locations of files and headers for gcc, and changes to how executables are built. As a result, the PGI compiler install process has to adjust with each release, and the code shipped needs to be updated or expanded. Most compiler releases will be compatible with previous Linux versions existing at that time, and we make an effort to be compatible with future releases, by trying to use the most current versions of libraries and headers from the installation for unfamiliar releases.

Here are the PGI compiler releases and the distributions of Linux they work with. Note, as is commonly done, we only update the most current release for a new distribution of Linux. The Linux distribtuion involved comes with a specific glibc and gcc version. If you use the versions that are bundled as part of the development tools for that Linux release, and they match another supported release below, then your installation will probably succeed. Not all versions of gcc/glibc can work with any Linux release/kernel, however, so any custom combination of Linux with gcc tools and libs should not be too different from the bundled release.

Here is a table of the Releases available and how to bring them up to their most up-to-date version. Be sure to add any patches after the latest version is downloaded and installed.

Release Linux Releases Verified to Install Newest Version Notes
2017 64-bit: RHEL 5.3–7.3, SLES 11/12 SP2,
openSUSE 11.1–Leap 42.2, Fedora 10–24, CentOS 5.11–7.1, Ubuntu 10.04–16.10
17.4 Release & version info
2016 32-bit and 64-bit: RHEL 5.3–7.2, SUSE 10.2–11.4, SLES 11/12 SP1,
openSUSE 11.1–Leap 42.1, Fedora 10–23, CentOS 5.11–7.1, Ubuntu 10.04–15.10
16.10 Release & version info
2015 32-bit and 64-bit: Red Hat 9.0, and RHEL 4.8–7.0, SUSE 9.3–11.2, SLES 10/11 SP3,
openSUSE 10.2–13.2, Fedora 4–21, CentOS 5.2–7, Ubuntu 8.04–14.10
15.10 Release & version info
2014 32-bit and 64-bit: Red Hat 9.0, and RHEL 4.0/5.0/5.3/6.0/6.2/6.3/6.5,
SUSE 9.0/9.1/9.2/9.3/10.0/10.1/10.2/10.3/11.0/11.1/11.2,
SLES 9/10/11 SP1 & SP2, openSUSE 12.2, Fedora 4–20, CentOS 6.4,
Ubuntu 8/9/9.04/10.04/10.10/11.04/11.11/12.04/12.10/13.04/13.10
14.10 Release & version info
2013 32-bit and 64-bit: Red Hat 9.0, and RHEL 4.0/5.0/5.3/6.0/6.2/6.3,
SUSE 9.0/9.1/9.2/9.3/10.0/10.1/10.2/10.3/11.0/11.1/11.2,
SLES 9/10/11 SP1 & SP2, openSUSE 12.2, Fedora 4–17
Ubuntu 8/9/9.04/10.04/10.10/11.04/11.11/12.04/12.10
13.10 Release & version info
2012 32-bit and 64-bit: Red Hat 9.0, and RHEL 4.0/5.0/5.3/6.0/6.2,
SuSE 9.0/9.1/9.2/9.3/10.0/10.1/10.2/10.3/11.0/11.1,
SLES 9/10/11 SP1, Fedora 4–16,
Ubuntu 8/9/9.04/10.04/10.10/11.04/11.11
12.10 Release & version info
2011 32-bit and 64-bit: Red Hat 9.0, and RHEL 4.0/5.0/5.3/6.0,
SuSE 9.0/9.1/9.2/9.3/10.0/10.1/10.2/10.3/11.0/11.1,
SLES 9/10/11, Fedora 4/5/6/7/8/9/10/11/12/13/14,
Ubuntu 8/9/9.04/10.04/10.10/11.04
11.10 Release & version info
2010 32-bit and 64-bit: Red Hat 9.0, and RHEL 4.0/5.0/5.3,
SuSE 9.0/9.1/9.2/9.3/10.0/10.1/10.2/10.3/11.0/11.1,
SLES 9/10/11, Fedora 4/5/6/7/8/9/10/11, Ubuntu 8/9/9.04
10.9 Release & version info
9.0 32-bit: Red Hat 9.0, and RHEL 3.0/4.0/5.0/5.3,
SuSE 9.0/9.1/9.2/9.3/10.0/10.1/10.2/10.3,
SLES 9/10/11, Fedora 4/5/6/7/8/9/10/11, Ubuntu 8/9
64-bit: SLES 9/10/11, SuSE 9.0/9.1/9.2/9.3/10.0/10.1/10.2/10.3/11.0/11.1,
RHEL 3.0/4.0/5.0/5.3, Fedora 4/5/6/7/8/9/10/11, Ubuntu 8/9
9.0-4 Release info
8.0 32-bit: Red Hat 9.0, and RHEL 3.0/4.0/5.0/5.3,
SuSE 9.0/9.1/9.2/9.3/10.0/10.1/10.2/10.3,
SLES 9/10/11, Fedora 4/5/6/7/8/9, Ubuntu 8
64-bit: SLES 9/10/11, SuSE 9.0/9.1/9.2/9.3/10.0/10.1/10.2/10.3,
RHEL 3.0/4.0/5.0/5.3, Fedora 4/5/6/7/8/9, Ubuntu 8
8.0-6 Release info
7.2 32-bit: Red Hat 9.0, and RHEL 3.0/4.0/5.0,
SuSE 9.0/9.1/9.2/9.3/10.0/10.1/10.2/10.3,
SLES 9/10, Fedora Core 4/5/6/7/8
64-bit: SLES 9/10, SuSE 9.0/9.1/9.2/9.3/10.0/10.1/10.2/10.3,
RHEL 3.0/4.0/5.0, Fedora Core 4/5/6/7/8
7.2-5 Release info


Which releases of PGI compilers work with each Windows version?

PGI Professional Edition for Windows. Versions are added as they become available.

Release Windows Versions Verified to Install Newest Version Notes
2017 64-bit: 7/8/8.1/10/Server 2008 R2/Server 2012/Server 2016 17.4 Release & version info
2016 32-bit and 64-bit: 7/8/8.1/10/Server 2008 R2/Server 2012 16.10 Release & version info
2015 32-bit and 64-bit: 7/8/8.1/Server 2008 R2/Server 2012 15.10 Release & version info
2014 32-bit: XP/Vista/7/8/8.1/Server 2003/Server 2008/Server 2012
64-bit: XP x64/Vista/7/8/8.1/Server 2003 x64/Server 2008 R2/Server 2012
14.10 Release & version info
2013 32-bit: XP/Vista/7/8/Server 2003/Server 2008/Server 2012
64-bit: XP x64/Vista/7/8/Server 2003 x64/Server 2008 R2/Server 2012
13.10 Release & version info
2012 32-bit: XP/Vista/7/Server 2003
64-bit: XP x64/Vista/7/Server 2003 x64/Compute Cluster Server/
HPC Server 2008 R2
12.10 Release & version info
2011 32-bit: XP/Vista/7*/Server 2003
64-bit: XP x64/Vista/Server 2003 x64/Compute Cluster Server/
HPC Server 2008 R2
11.10 Release & version info
2010 32-bit: XP/Vista/7*/Server 2003
64-bit: XP x64/Vista/Server 2003 x64/Compute Cluster Server/
HPC Server 2008
10.9 Release & version info
9.0 32-bit: XP/Vista/Server 2003
64-bit: XP x64/Vista/Server 2003 x64/Compute Cluster Server/HPC Server 2008
9.0-4 Release info
8.0 32-bit: XP/Vista/Server 2003
64-bit: XP x64/Vista/Server 2003 x64/Compute Cluster Server/HPC Server 2008
8.0-6 Release info
7.2 32-bit: 2003
64-bit: XP x64/Vista/Server 2003 x64/Compute Cluster Server
7.2-5 Release info
*Note on Windows 7 compatibility: FLEXlm versions delivered with PGI 2010 and PGI 2011 releases were compatible with only the 32-bit versions of Windows 7. PGI 2012 and newer releases are compatible with both the 32-bit and the 64-bit versions of Windows 7.


Which releases of PGI Visual Fortran compilers work with each Windows version?

Versions are added as they become available.

Note: PVF is NOT compatible with Visual Studio Express. PVF will not install or run successfully on systems with VS Express present regardless of the VS Express version or PVF version.

Release Windows Versions Verified to Install Newest Version Notes
2017 64-bit: 7/8/8.1/10/Server 2008 R2/Server 2012/Server 2016
Supports Microsoft Visual Studio 2015 (VS 2015 shell included), 2013
17.4 Release & version info
2016 32-bit and 64-bit: 7/8/8.1/10/Server 2008 R2/Server 2012
Supports Microsoft Visual Studio 2015 (VS 2015 shell included), 2013
16.10 Release & version info
2015 32-bit and 64-bit: 7/8/8.1/Server 2008 R2/Server 2012
Supports Microsoft Visual Studio 2013 (VS 2013 shell included), 2012, 2010 and 2008
15.10 Release & version info
2014 32-bit: XP/Vista/7/8/8.1/Server 2003/HPC Server 2008 R2/Server 2012
64-bit: XP x64/Vista/7/8/Server 2003 x64/HPC Server 2008 R2/Server 2012
Supports Microsoft Visual Studio 2013 (VS 2013 shell included), 2012, 2010 and 2008
14.10 Release & version info
2013 32-bit: XP/Vista/7/8/Server 2003
64-bit: XP x64/Vista/7/8/Server 2003 x64/Compute Cluster Server/HPC Server 2008 R2
Supports Microsoft Visual Studio 2012 (VS 2012 shell included), 2010 and 2008
13.10 Release & version info
2012 32-bit: XP/Vista/7/Server 2003
64-bit: XP x64/Server 2003 x64/Compute Cluster Server/Vista/7/HPC Server 2008 R2
Supports Microsoft Visual Studio 2010 (VS 2010 shell included) and 2008
12.10 Release & version info
2011 32-bit: XP/Vista/7/Server 2003
64-bit: XP x64/Server 2003 x64/Compute Cluster Server/Vista/HPC Server 2008 R2
Supports Microsoft Visual Studio 2010 (VS 2010 shell included) and 2008
11.10 Release & version info
2010 32-bit: XP/Vista/7/Server 2003
64-bit: XP x64/Server 2003 x64/Compute Cluster Server/Vista/HPC Server 2008
Supports Microsoft Visual Studio 2010, 2008 (VS 2008 shell included) and 2005
10.9 Release & version info
9.0 32-bit: XP/Vista/Server 2003
64-bit: XP x64/Server 2003 x64/Compute Cluster Server/Vista/HPC Server 2008
Supports Microsoft Visual Studio 2008 (VS 2008 shell included) and 2005
9.0-4 Release info
8.0 32-bit: XP/Vista/Server 2003
64-bit: XP x64/Server 2003 x64/Compute Cluster Server/Vista/HPC Server 2008
Supports Microsoft Visual Studio 2008 (VS 2008 shell included) and 2005
8.0-6 Release info
7.2 32-bit: XP/Vista/Server 2003
64-bit: XP x64/Server 2003 x64/Compute Cluster Server/Vista
Supports Microsoft Visual Studio 2008 and 2005
7.2-5 Release info
See the note above on Windows 7 compatibility.

Can I install PVF on a platform with Visual Studio Express installed?

No. PVF will not install and run successfully on a system with Visual Studio Express installed. Even installing PVF with the VS 2010 shell onto a system with VS Express 2008 already installed will not work. You must first remove VS Express before installing PVF.


Which releases of PGI compilers work with each macOS version on Intel-based Macs?

PGI products for macOS. Versions are added as they become available.
Note: PGI compilers and tools are supported only on Intel processor-based Macintosh computers.

Release macOS Versions Verified to Install Newest Version Notes
2017 64-bit: 10.9 Mavericks/ 10.10 Yosemite/10.11 El Capitan
10.12 Sierra
17.4 Release & version info
2016 32-bit and 64-bit: 10.8 Mountain Lion/10.9 Mavericks/ 10.10 Yosemite/
10.11 El Capitan
16.10 Release & version info
2015 32-bit and 64-bit: OS X 10.7 Lion/10.8 Mountain Lion/10.9 Mavericks/
10.10 Yosemite
15.10 Release & version info
2014 32-bit: OS X 10.6 Snow Leopard/10.7 Lion/10.8 Mountain Lion/10.9 Mavericks
64-bit: OS X 10.6 Snow Leopard/10.7 Lion/10.8 Mountain Lion/10.9 Mavericks
14.10 Release & version info
2013 32-bit: OS X 10.6 Snow Leopard/10.7 Lion/10.8 Mountain Lion
64-bit: OS X 10.6 Snow Leopard/10.7 Lion/10.8 Mountain Lion
13.10 Release & version info
2012 32-bit: Mac OS X 10.6 Snow Leopard/10.7 Lion
64-bit: Mac OS X 10.6 Snow Leopard/10.7 Lion
12.10 Release & version info
2011 32-bit: Mac OS X 10.5 Leopard/10.6 Snow Leopard/10.7 Lion
64-bit: Mac OS X 10.5 Leopard/10.6 Snow Leopard/10.7 Lion
11.10 Release & version info
2010 32-bit: Mac OS X 10.5 Leopard/10.6 Snow Leopard
64-bit: Mac OS X 10.5 Leopard/10.6 Snow Leopard
10.9 Release & version info
9.0 32-bit: Mac OS X 10.4.9 Tiger/10.5 Leopard/10.6 Snow Leopard
64-bit: Mac OS X 10.4.9 Tiger/10.5 Leopard
9.0-4 Release info
8.0 32-bit: Mac OS X 10.4.9 Tiger/10.5 Leopard
64-bit: Mac OS X 10.4.9 Tiger/10.5 Leopard
8.0-6 Release info
7.2 32-bit: Mac OS X 10.4.9 Tiger/10.5 Leopard
64-bit: Mac OS X 10.4.9 Tiger/10.5 Leopard
7.2-5 Release info


Which versions of PGI Accelerator compilers support which CUDA Toolkit versions?

Applies to all PGI Accelertor products. New PGI versions are added when the CUDA Toolkit is updated.
Note: Default toolkit is displayed in bold.

Version Supported CUDA Toolkit Versions Support Operating Systems Release Notes
17.1 7.5, 8.0 64-bit Linux. 64-bit Windows 7 or newer. 64-bit macOS 10.9 or newer. CUDA 8.0
16.1 7.0, 7.5 64-bit Linux. 64-bit Windows 7 or newer. 64-bit OS X 10.9 or newer. CUDA 7.5
15.9 6.5, 7.0, 7.5 64-bit Linux. 64-bit Windows 7 or newer. 64-bit OS X 10.9 or newer.
32-bit Windows execution.
CUDA 7.0
15.4 5.5, 6.5, 7.0 64-bit Linux. 64-bit and 32-bit Windows 7 and newer.
64-bit OS X 10.8 and newer.
CUDA 6.5
15.1 5.5, 6.0, 6.5 64-bit and 32-bit Linux. 64-bit and 32-bit Windows XP and newer.
64-bit and 32-bit OS X 10.8 and newer.
14.9 6.0, 6.5 64-bit and 32-bit Linux. 64-bit and 32-bit Windows XP and newer.
64-bit and 32-bit OS X 10.8 and newer.
CUDA 6.0
14.4 5.5, 6.0 64-bit and 32-bit Linux.
64-bit and 32-bit Windows XP and newer; Server 2008 R2 and newer.
64-bit and 32-bit OS X 10.7.5 and newer.
CUDA 5.5
14.4 5.0, 5.5 64-bit and 32-bit Linux.
64-bit and 32-bit Windows XP and newer; Server 2008 R2 and newer.
64-bit and 32-bit OS X 10.7 and newer.
CUDA 5.0


How do I uninstall PGI software?

To remove a PGI installation from a Linux platform, simply remove the $PGI directory.

% cd /
% chmod -R a+w  $PGI; rm -rf $PGI

To remove a particular release from a $PGI directory where multiple PGI releases are installed, go down another directory level, and remove the desired installation directory. This example removes the 64-bit 16.5 installation.

% cd $PGI/linux86-64
% ls
16.5/  16.7/  16.9/ 16.10/ 2016/
% chmod -R a+w ./16.5; rm -rf 16.5

To remove a PGI installation from a macOS system, simply remove the /opt/pgi directory.

% cd /opt
% rm -rf pgi

To remove Free PGI for OS X, simply drag the icon from the Applications folder to the Trash.

To remove a PGI installation from a Windows platform, use the Control Panel-> Programs and Features facility, and uninstall the PGI product or version in question.

Community Edition

A no-cost license to a recent release of the PGI Fortran, C and C++ compilers and tools for x86-64 and OpenPOWER multicore CPUs and NVIDIA Tesla GPUs, including all OpenACC, OpenMP and CUDA Fortran features. The PGI Community Edition enables development of performance-portable HPC applications with uniform source code across the most widely used parallel processors and systems.

Download now

Community Edition

Includes all features of the PGI Community Edition, plus a perpetual license to current and all previous releases of the PGI compilers and tools. Offers technical support with frequent PGI updates that include the latest PGI feature enhancements, performance improvements and bug fixes. The PGI Professional Edition is for HPC experts who need cutting edge compilers and support for production software development.

Learn more

Click me