LazyBatching: An SLA-aware Batching System for Cloud Machine Learning Inference

TL;DR

LazyBatching is a novel SLA-aware batching system for cloud ML inference that improves response time, throughput, and SLA satisfaction by batching at the node level rather than entire graphs, effectively handling dynamic traffic.

Contribution

It introduces a node-level batching approach that considers SLA constraints, outperforming traditional graph batching in cloud ML inference systems.

Findings

15x improvement in response time

1.5x increase in throughput

5.5x better SLA satisfaction

Abstract

In cloud ML inference systems, batching is an essential technique to increase throughput which helps optimize total-cost-of-ownership. Prior graph batching combines the individual DNN graphs into a single one, allowing multiple inputs to be concurrently executed in parallel. We observe that the coarse-grained graph batching becomes suboptimal in effectively handling the dynamic inference request traffic, leaving significant performance left on the table. This paper proposes LazyBatching, an SLA-aware batching system that considers both scheduling and batching in the granularity of individual graph nodes, rather than the entire graph for flexible batching. We show that LazyBatching can intelligently determine the set of nodes that can be efficiently batched together, achieving an average 15x, 1.5x, and 5.5x improvement than graph batching in terms of average response time, throughput, and SLA satisfaction, respectively.