Nvidia company, Cray company, PGI, and capital launched OpenACC "parallel computation of programming standards
A group including NVIDIA, clay, Portland group (PGI) and electronics wholesale CAPS of the company has announced that the enterprise OpenACC known parallel programming standards. OpenACC is an open parallel programming standard design, make millions of science and technology of the programmers are easy to use of CPU/GPU heterogeneous calculation system advantage. It allows parallel program to provide simple tips, known as the "instructions", the compiler, identification code acceleration, without the need to adapt to the bottom or programmers modify the code itself. Through the expose parallel compiler, command allows the compiler mapped to the accelerator calculation do meticulous work. OpenACC purpose is in a programmer in chemistry, biology, physics, and data analysis, the weather and climate, intelligence, and many other areas of work widely. From the existing Cray company, PGI and CAPS the compiler is expected to OpenACC standards in the first quarter of 2012 to start to provide preliminary support. The standard is fully compatible and interoperability NVIDIA CUDA parallel programming framework, its purpose is permitted in the most performance optimization of the accelerator detailed control.