On Optimizing Complex Stencils on GPUs

Citations Over TimeTop 10% of 2019 papers

Abstract

Stencil computations are often the compute-intensive kernel in many scientific applications. With the increasing demand for computational accuracy, and the emergence of massively data-parallel high-bandwidth architectures like GPUs, stencils have steadily become more complex in terms of the stencil order, data accesses, and reuse patterns. Many prior efforts have focused on optimizing simpler stencil computations on various platforms. However, existing stencil code generators face challenges in optimizing such complex multi-statement stencil DAGs. This paper addresses the challenges in optimizing high-order stencil DAGs on GPUs by focusing on two key considerations: (1) enabling the domain expert to guide the code optimization, which may otherwise be extremely challenging for complex stencils; and (2) using bottleneck analysis via runtime profiling to guide the application of optimizations, and the tuning of various code generation parameters. We implement these abstractions in a prototype code generation framework termed ARTEMIS, and evaluate its efficacy over multiple stencil kernels with varying complexity and operational intensity on an NVIDIA P100 GPU.

Related Papers

• → Using GPUs for Parallel Stencil Computations in Relativistic Hydrodynamic Simulation(2014)7 cited
• → Electromagnetic scattering simulations using overset grids on massively parallel computing platforms(2002)1 cited
• → Large-Scale Gas Turbine Simulations on GPU Clusters(2011)
• → Studies on the Efficiency of Massively Parallel Computational Method for Large-Scale Computation based on Unstructured Grid(1996)
• → M3E_LINSOL: A Scalable Linear Solver Package for Reservoir Simulations on Massively Parallel Computers(2017)