It is not that computing cores aren't getting faster. Instead, processors are getting more parallel, which is a trend that is likely to continue. To harness advances in parallel computing, NVIDIA and its partners developed the OpenACC standard, which NVIDIA says "simplifies parallel programming for modern processors, like GPUs". In order to simplify access to OpenACC for researchers, NVIDIA has released the new NVIDIA OpenACC Toolkit, a free, all-in-one suite of OpenACC parallel programming tools. NVIDIA claims that scientists can do "more science, less programming" from the solution, which features "the industry-leading" PGI Accelerator Fortran/C Workstation Compiler Suite for Linux. The compiler is free to academic developers and researchers. The toolkit also includes the NVProf Profiler, which gives guidance on where to add OpenACC "directives"—that is, simple compiler hints to accelerate code, as well as simple, real-world code samples.
