Trivance: Latency-Optimal AllReduce by Shortcutting Multiport Networks

TL;DR

Trivance is a new AllReduce algorithm that reduces communication steps and congestion, achieving latency-optimal performance in large-scale distributed systems with bidirectional ring and multidimensional torus topologies.

Contribution

It introduces Trivance, an AllReduce algorithm that completes in log_3 n steps, triples communication distance per step, and reduces congestion while maintaining bandwidth optimality.

Findings

Improves state-of-the-art by 5-30% for message sizes up to 8 MiB.

Reduces congestion by a factor of three compared to Bruck's algorithm.

Maintains latency advantage in multidimensional torus networks.

Abstract

AllReduce is a fundamental collective operation in distributed computing and a key performance bottleneck for large-scale training and inference. Its completion time is determined by the number of communication steps, which dominates latency-sensitive workloads, and the communication distance affecting both latency- and bandwidth-bound regimes. Direct-connect topologies, such as torus networks used in Google's TPUv4, are particularly prone to large communication distances due to limited bisection bandwidth. Latency-optimal algorithms such as Bruck's complete AllReduce in $\log_3 n$ steps on a bidirectional ring, but incur large communication distances that result in substantial congestion. In contrast, recent approaches such as Swing reduce communication distance and congestion, but are inherently required to perform $\log_2 n$ steps to complete AllReduce, sacrificing latency-optimality. In this paper, we present Trivance, a novel AllReduce algorithm that completes within $\log_3 n$ steps, while reducing congestion compared to Bruck's algorithm by a factor of three and preserving bandwidth-optimality. Trivance exploits both transmission ports of a bidirectional ring within each step to triple the communication distance along both directions simultaneously. Furthermore, by performing joint reductions, Trivance improves both the number of steps and network congestion. We further show that Trivance extends naturally to multidimensional torus networks, retaining its latency advantage while achieving performance comparable to bandwidth-optimal algorithms for large messages. Our empirical evaluation shows that Trivance improves state-of-the-art approaches by 5-30% for message sizes up to 8\,MiB, in high-bandwidth settings up to 32MiB and for 3D tori up to 128MiB. Throughout the evaluation, Trivance remains the best-performing latency-optimal algorithm.