Nvidia Adds Support For OpenACC On x86 Processors
Nvidia, the tech company known for its GPUs and SOCs used for mobile computing, has added support for OpenACC parallel programming on x86 architecture processors, all possible through the recently launched v15.10 PGI Accelerator Fortran, C and C++ compilers.
According to the company the compilers will bring improved performance as well as portability so that OpenACC enabled source code can be executed on multi core CPUs or a GPU accelerator in parallel running.
OpenACC was created in 2011 by a group of companies comprising Nvidia, Cray, the Portland Group (PGI) and CAPS enterprise. The open parallel programming standard is designed to enable the millions of scientific and technical programmers to easily take advantage of heterogeneous CPU/GPU computing systems.
OpenACC enables programmers to give out directives to the compiler, highlighting which areas of the code need to be accelerated without requiring any modification from the programmer.
“This capability provides tremendous flexibility for programmers, enabling them to develop applications that can take advantage of multiple system architectures with a single version of their source code,” explained Nvidia.
According to Nvidia, another major benefit for running OpenACC on multicore CPUs is the common programming model across a variety of CPUs and GPUs in Fortran, C and C++ allowing for faster exploitation in a program as well as scalable performance across multicore setups.
The OpenACC Toolkit from Nvidia can be downloaded from here.
Source: The Inquirer