The Portland Group is now shipping the PGI Cuda C and C++ compilers for systems based on the industry standard general-purpose 64-bit and 32-bit x86 architectures. The new compilers enable more developers to use the Cuda parallel programming model to optimise the performance of the critical parts of their code base, targeting servers and clusters with or without Nvidia GPUs. When run on x86-based systems, PGI Cuda C/C++ applications perform parallel execution by using the multiple processor cores and Streaming SIMD (Single Instruction Multiple Data) Extensions (SSE), including the new AVX instructions available on the latest generation of x86 compatible CPUs from Intel and AMD.
PGI will roll out the x86 Cuda C/C++ compilers in three phases. Phase one, available now, demonstrates the capabilities of the technology and allows developers to begin working with the compilers. Phase two, scheduled for the fourth quarter of 2011, will include performance optimisations intended to extract maximum performance of Cuda programs running on the x86 target platform. Phase three planned for mid 2012, will include support for PGI Unified Binary technology - the ability to run one executable on both CPUs and GPUs. The compilers for x86 are included as part of the PGI Accelerator product line.
All PGI Accelerator products support both Cuda programming and the high-level directive-based programming model targets scientific and engineering domain experts working in high-performance computing. PGI Accelerator compilers are currently available for C99 and Fortran 2003, and products are supported on the Linux, Apple Mac OS X and Microsoft Windows operating systems.