Reliable Performance

PGI compilers consistently rate among the highest on high-performance computing industry benchmarks. PGI compilers are now over 42% faster than GCC in floating point performance and over 20% faster on integer. Using PGI compilers is like adding more cluster nodes for a fraction of the cost.

Easy MPI Debugging

PGDBG^® is a powerful yet easy-to-use graphical debugger for 64-bit MPI and OpenMP parallel applications. Invest five minutes now watching the PGDBG demo and you could save hours the next time you need to debug your parallel code.

Strategic Independence

PGI Unified Binary™ technology lets you build a single x64 binary executable optimized for both AMD and Intel processors. Find your road to strategic independence. Maximize flexibility and eliminate the need to target and optimize for separate processors.

Multi-core Answers

The first step towards maximizing multi-core performance is optimizing single core performance. PGPROF^® can help. PGPROF is an interactive OpenMP and MPI parallel performance profiler that can help you quickly see where where your applications is spending time.

Peak Performance

With PGI, you get a comprehensive set of tools and technologies to help you squeeze peak performance from your code. Optimizing compilers, parallel performance profiling with scalability analysis, and hand-tuned BLAS, LAPACK and FFT numerical libraries to name just a few. Compare prices and features.

HPC Your Way

PGI products support multi-core 64-bit AMD Opteron and Intel Xeon processors running over 30 popular versions of Linux and Windows. PGI MPI tools support MPICH, MPICH-2 and MSMPI over Ethernet and MVAPICH over InfiniBand. If your organization relies on a variety of HPC cluster technologies, check out the PGI CDK Cluster Development Kit.