PGI has a long history of innovation in HPC. The following list summarizes a number of key PGI technologies in use today.

PGI Unified Binary™ technology is the ability to generate a single executable file containing code sequences optimized for multiple CPUs and accelerators.

PGI Accelerator™ compilers allow programmers to accelerate applications on CPU+accelerator platforms by adding OpenACC compiler directives to existing high-level standard-compliant Fortran, C and C++ programs and then recompiling with appropriate compiler options.

Common Compiler Feedback Format (CCFF) is PGI's open specification that defines the format for storing the wealth of information about what the compiler did and could not do to optimize the code during compilation. Using CCFF, HPC tools providers can enhance their products to offer more and better information about optimizing performance.

CUDA Fortran is a Fortran analog to NVIDIA CUDA C. CUDA Fortran includes a Fortran 2003 compiler and tool chain for programming NVIDIA GPUs using Fortran. CUDA Fortran is supported on Linux, macOS and Windows.

CUDA-x86 is PGI's CUDA C/C++ compiler for x86 platforms. It lets developers using CUDA compile and optimize their CUDA applications to run on x86-based workstations, servers and clusters with or without an NVIDIA GPU accelerators.

Program Analysis Summary Output During compiling, PGI compilers have the ability to gather a detailed cross-file summary of a program’s internal structure, including function call sites, variable declarations, reads, or modifications, OpenACC and OpenMP regions, and more.

Click me