An In-Depth Analysis of the Slingshot Interconnect

TL;DR

This paper presents Slingshot, a high-radix interconnection network designed for large-scale systems, offering low latency, adaptive routing, and compatibility with Ethernet, improving application performance and congestion management.

Contribution

It introduces Slingshot, a novel interconnect with high-radix switches, optimized Ethernet protocol, and advanced routing, enabling scalable, high-performance datacenter and HPC applications.

Findings

Applications on Slingshot experience less congestion impact.

Slingshot achieves high scalability with at most three switch hops.

It provides efficient adaptive routing and congestion control.

Abstract

The interconnect is one of the most critical components in large scale computing systems, and its impact on the performance of applications is going to increase with the system size. In this paper, we will describe Slingshot, an interconnection network for large scale computing systems. Slingshot is based on high-radix switches, which allow building exascale and hyperscale datacenters networks with at most three switch-to-switch hops. Moreover, Slingshot provides efficient adaptive routing and congestion control algorithms, and highly tunable traffic classes. Slingshot uses an optimized Ethernet protocol, which allows it to be interoperable with standard Ethernet devices while providing high performance to HPC applications. We analyze the extent to which Slingshot provides these features, evaluating it on microbenchmarks and on several applications from the datacenter and AI worlds, as well as on HPC applications. We find that applications running on Slingshot are less affected by congestion compared to previous generation networks.