ChASE -- A Distributed Hybrid CPU-GPU Eigensolver for Large-scale Hermitian Eigenvalue Problems

TL;DR

ChASE is a scalable distributed hybrid CPU-GPU eigensolver designed for large Hermitian eigenproblems, leveraging Chebyshev polynomial filters to efficiently compute partial eigenpairs on modern supercomputers.

Contribution

This work extends ChASE to support distributed hybrid CPU-multi-GPU architectures, demonstrating high scalability and performance on large dense eigenproblems.

Findings

Achieves good scaling up to 144 nodes with 526 GPUs

Handles dense eigenproblems up to size 360,000

Outperforms traditional solvers in parallel efficiency

Abstract

As modern massively parallel clusters are getting larger with beefier compute nodes, traditional parallel eigensolvers, such as direct solvers, struggle keeping the pace with the hardware evolution and being able to scale efficiently due to additional layers of communication and synchronization. This difficulty is especially important when porting traditional libraries to heterogeneous computing architectures equipped with accelerators, such as Graphics Processing Unit (GPU). Recently, there have been significant scientific contributions to the development of filter-based subspace eigensolver to compute partial eigenspectrum. The simpler structure of these type of algorithms makes for them easier to avoid the communication and synchronization bottlenecks typical of direct solvers. The Chebyshev Accelerated Subspace Eigensolver (ChASE) is a modern subspace eigensolver to compute partial extremal eigenpairs of large-scale Hermitian eigenproblems with the acceleration of a filter based on Chebyshev polynomials. In this work, we extend our previous work on ChASE by adding support for distributed hybrid CPU-multi-GPU computing architectures. Our tests show that ChASE achieves very good scaling performance up to 144 nodes with 526 NVIDIA A100 GPUs in total on dense eigenproblems of size up to $360$k.