BOAST: A metaprogramming framework to produce portable and efficient computing kernels for HPC applications

Brice Videau (1, 2, 3, 4) , Kevin Pouget (5, 2, 3) , Luigi Genovese (4) , Thierry Deutsch (4) , Dimitri Komatitsch (6) , Frédéric Desprez (2, 3) , Jean-François Méhaut (5, 2, 3)

1 CNRS - Centre National de la Recherche Scientifique
2 LIG - Laboratoire d'Informatique de Grenoble
3 CORSE - Compiler Optimization and Run-time Systems
4 LSIM - Laboratory of Atomistic Simulation
5 UGA [2016-2019] - Université Grenoble Alpes [2016-2019]
6 LMA - Laboratoire de Mécanique et d'Acoustique [Marseille]

The portability of real high-performance computing (HPC) applications on new platforms is an open and very delicate problem. Especially, the performance portability of the underlying computing kernels is problematic as they need to be tuned for each and every platform the application encounters. This article presents BOAST, a metaprogramming framework dedicated to computing kernels. BOAST allows the description of a kernel and its possible optimizations using a domain-specific language. BOAST runtime will then compare the different versions’performance as well as verify their exactness. BOAST is applied to three use cases: a Laplace kernel in OpenCL and two HPC applications BigDFT (electronic density computation) and SPECFEM3D (seismic and wave propagation).

Mots clés: Portability; Genericity; Code generation; Nonregression; Testing; Productivity and software design; Autotuning; High-performance computing

Domaines: Acoustique [physics.class-ph]; Calcul parallèle, distribué et partagé [cs.DC]; Géophysique [physics.geo-ph]

Brice Videau, Kevin Pouget, Luigi Genovese, Thierry Deutsch, Dimitri Komatitsch, et al.. BOAST: A metaprogramming framework to produce portable and efficient computing kernels for HPC applications. International Journal of High Performance Computing Applications, 2018, 32 (1), pp.28-44. ⟨10.1177/1094342017718068⟩. ⟨hal-01620778⟩