Yet Another Tensor Toolbox for discontinuous Galerkin methods and other applications

03/27/2019
by   Carsten Uphoff, et al.
0

The numerical solution of partial differential equations is at the heart of many grand challenges in supercomputing. Solvers based on high-order discontinuous Galerkin (DG) discretisation have been shown to scale on large supercomputers with excellent performance and efficiency, if the implementation exploits all levels of parallelism and is tailored to the specific architecture. However, every year new supercomputers emerge and the list of hardware-specific considerations grows, simultaneously with the list of desired features in a DG code. Thus we believe that a sustainable DG code needs an abstraction layer to implement the numerical scheme in a suitable language. We explore the possibility to abstract the numerical scheme as small tensor operations, describe them in a domain-specific language (DSL) resembling the Einstein notation, and to map them to existing code generators which generate small matrix matrix multiplication routines. The compiler for our DSL implements classic optimisations that are used for large tensor contractions, and we present novel optimisation techniques such as equivalent sparsity patterns and optimal index permutations for temporary tensors. Our application examples, which include the earthquake simulation software SeisSol, show that the generated kernels achieve over 50 considerably simplifies the implementation.

READ FULL TEXT
research
03/21/2023

Domain-specific implementation of high order Discontinuous Galerkin methods in spherical geometry

We assess two domain-specific languages included in the GridTools ecosys...
research
03/28/2020

Vectorization and Minimization of Memory Footprint for Linear High-Order Discontinuous Galerkin Schemes

We present a sequence of optimizations to the performance-critical compu...
research
11/29/2017

High-performance Implementation of Matrix-free High-order Discontinuous Galerkin Methods

Achieving a substantial part of peak performance on todays and future hi...
research
07/02/2019

A hp-adaptive discontinuous Galerkin solver for elliptic equations in numerical relativity

A considerable amount of attention has been given to discontinuous Galer...
research
04/18/2019

Towards whole program generation of quadrature-free discontinuous Galerkin methods for the shallow water equations

The shallow water equations (SWE) are a commonly used model to study tsu...
research
05/10/2017

TSFC: a structure-preserving form compiler

A form compiler takes a high-level description of the weak form of parti...
research
01/26/2018

Modeling of languages for tensor manipulation

Numerical applications and, more recently, machine learning applications...

Please sign up or login with your details

Forgot password? Click here to reset