white paper

Using OpenACC to exploit application parallelism with minimal effort

Exploiting parallelism for accelerating computation, is challenging and requires abstract thought

Exploiting Parallelism for High Performance Computing

This paper overviews the use of OpenACC (Open ACCelerators) to accelerate programs implementing high-performance compute algorithms such as machine learning applications. Using OpenACC directives enables a programmer to develop efficient programs with minimal effort that perform effectively on a number of architectures.

Share