Session

DescriptionIn this minisymposium we will discuss some of the efforts that are made in the area of improving performance portability and programming productivity for motif-based high-performance algorithms that are used in wide range of scientific applications. We will describe domain specific libraries (DSLs) that express mathematical/programming motifs (data objects and operations on those data objects), along with software back-ends that translate the library calls into high-performance code. By the use of a motif-aware software stack, the scientific application code written is much smaller than fully optimized code, with the applications-level code remaining unchanged in moving between platforms, thus leading to a less expensive development process. There will be four talks given in this minisymposium covering multiple motifs like structured grids, Fast Fourier Transforms (FFTs), particle methods and dense/sparse linear algebra, and different approaches to supporting motif-based DSLs. This minisymposium aims to bring together different groups focused on developing various motif-based high-performance codes using different software tools and automate the process as much as possible.