NVIDIA has launched the NVIDIA CUDA 6.5 production release, the latest version of its parallel computing platform and programming model. The CUDA toolkit brings GPU (graphics processing units)-accelerated computing to 64-bit ARM platforms. It enables a newly expanded range of efficient, HPC (high-performance computing) options to accelerate compute- intensive HPC and enterprise data center workloads, the company states.
CUDA 6.5 offers programmers with a platform to develop advanced scientific, engineering, mobile and HPC applications on GPU-accelerated ARM and x86 CPU-based systems.
Additional features include:
- Support for Microsoft Visual Studio 2013
- cuFFT callbacks capability that delivers higher performance custom processing on input or output data.
- Improved debugging for CUDA FORTRAN applications which offers Linux users debugging support for FORTRAN arrays, improved source-to-assembly code correlation and documentation.
- Application replay mode for faster analysis of complex scenarios using multiple hardware encounters.
- Updated CUDA Occupancy Calculator API (application programming interface) which frees programmers from having to manually configure kernel launches.
- New “nvprune” utility that prunes object files and libraries to only contain necessary device code for specified target architectures.
- Support for Block Sparse Row matrix format
For more information, visit NVIDIA.
Sources: Press materials received from the company and additional information gleaned from the company’s website.