HiDP: Hierarchical DNN Partitioning for Distributed Inference on Heterogeneous Edge Platforms

TL;DR

This work proposes a hierarchical DNN partitioning strategy (HiDP) for distributed inference on heterogeneous edge nodes that hierarchically partitions DNN workloads at both global and local levels by considering the core-level heterogeneity of edge nodes.

Abstract

Edge inference techniques partition and distribute Deep Neural Network (DNN) inference tasks among multiple edge nodes for low latency inference, without considering the core-level heterogeneity of edge nodes. Further, default DNN inference frameworks also do not fully utilize the resources of heterogeneous edge nodes, resulting in higher inference latency. In this work, we propose a hierarchical DNN partitioning strategy (HiDP) for distributed inference on heterogeneous edge nodes. Our strategy hierarchically partitions DNN workloads at both global and local levels by considering the core-level heterogeneity of edge nodes. We evaluated our proposed HiDP strategy against relevant distributed inference techniques over widely used DNN models on commercial edge devices. On average our strategy achieved 38% lower latency, 46% lower energy, and 56% higher throughput in comparison with other relevant approaches.

Figures (9)

• Figure 1: Inference latency of DNN models with different workload partitioning configurations.
• Figure 2: Comparison of global and hierarchical dnn partitioning strategies.
• Figure 3: Overview of the proposed HiDP framework. In this instance, the framework is run on device-1, partitioning the dnn model globally through model partitioning and locally through data partitioning.
• Figure 4: Workflow of the leader and follower nodes in the HiDP controller
• Figure 5: Inference (a). latency and (b). energy consumption of different strategies for targeted dnn workloads.