SIAM Journal on Scientific Computing, Volume 42, Issue 6, Page C384-C409, January 2020.
Parallel implementations of linear iterative solvers generally alternate between phases of data exchange and phases of local computation. Increasingly large problem sizes and more heterogeneous compute architectures make load balancing and the design of low latency network interconnects that are able to satisfy the communication requirements of linear solvers very challenging tasks. In particular, global communication patterns such as inner products become increasingly limiting at scale. We explore the use of asynchronous communication based on one-sided Message Passing Interface primitives in the context of domain decomposition solvers. In particular, a scalable asynchronous two-level Schwarz method is presented. We discuss practical issues encountered in the development of a scalable solver and show experimental results obtained on a state-of-the-art supercomputer system that illustrate the benefits of asynchronous solvers in load balanced as well as load imbalanced scenarios. Using the novel method, we can observe speedups of up to four times over its classical synchronous equivalent.

中文翻译：

---

可伸缩的异步域分解求解器

SIAM科学计算杂志，第42卷，第6期，第C384-C409页，2020年1月。
线性迭代求解器的并行实现通常在数据交换阶段和本地计算阶段之间交替。越来越大的问题规模和更加异构的计算体系结构使得负载平衡以及能够满足线性求解器通信要求的低延迟网络互连的设计非常具有挑战性。特别地，诸如内部产品之类的全球通信模式在规模上变得越来越受限制。我们探索在域分解求解器的上下文中基于单面消息传递接口原语的异步通信的使用。特别是，提出了一种可扩展的异步两级Schwarz方法。我们讨论了可伸缩求解器的开发过程中遇到的实际问题，并显示了在最新的超级计算机系统上获得的实验结果，这些结果说明了异步求解器在负载平衡和负载不平衡情况下的优势。使用新颖的方法，我们可以观察到其经典同步等效技术上的加速高达四倍。