SParSH-AMG: A library for hybrid CPU-GPU algebraic multigrid and preconditioned iterative methods

TL;DR

This paper introduces SParSH-AMG, a hybrid CPU-GPU algebraic multigrid library that efficiently uses both resources, reduces GPU memory needs, and enhances the performance of iterative solvers for large algebraic systems.

Contribution

The paper presents a novel hybrid CPU-GPU AMG framework that minimizes GPU memory usage while maintaining high performance, enabling the solution of larger systems.

Findings

Hybrid AMG achieves comparable performance to GPU-only methods.

Significantly lower GPU memory requirements for large systems.

Effective as a preconditioner with Krylov solvers.

Abstract

Hybrid CPU-GPU algorithms for Algebraic Multigrid methods (AMG) to efficiently utilize both CPU and GPU resources are presented. In particular, hybrid AMG framework focusing on minimal utilization of GPU memory with performance on par with GPU-only implementations is developed. The hybrid AMG framework can be tuned to operate at a significantly lower GPU-memory, consequently, enables to solve large algebraic systems. Combining the hybrid AMG framework as a preconditioner with Krylov Subspace solvers like Conjugate Gradient, BiCG methods provides a solver stack to solve a large class of problems. The performance of the proposed hybrid AMG framework is analysed for an array of matrices with different properties and size. Further, the performance of CPU-GPU algorithms are compared with the GPU-only implementations to illustrate the significantly lower memory requirements.