RankMap: Priority-Aware Multi-DNN Manager for Heterogeneous Embedded Devices

TL;DR

RankMap is introduced, a priority-aware manager specifically designed for multi-DNN tasks on heterogeneous embedded devices that addresses the extensive solution space of multi-DNN mapping through stochastic space exploration combined with a performance estimator.

Abstract

Modern edge data centers simultaneously handle multiple Deep Neural Networks (DNNs), leading to significant challenges in workload management. Thus, current management systems must leverage the architectural heterogeneity of new embedded systems to efficiently handle multi-DNN workloads. This paper introduces RankMap, a priority-aware manager specifically designed for multi-DNN tasks on heterogeneous embedded devices. RankMap addresses the extensive solution space of multi-DNN mapping through stochastic space exploration combined with a performance estimator. Experimental results show that RankMap achieves x3.6 higher average throughput compared to existing methods, while preventing DNN starvation under heavy workloads and improving the prioritization of specified DNNs by x57.5.

Figures (10)

• Figure 1: Average throughput $\mathcal{T}$ normalized by baseline.
• Figure 2: Potential Throughput $\mathcal{P} \forall$ DNNs.
• Figure 3: A high-level overview of RankMap.
• Figure 4: Reward calculation in RankMap by example.
• Figure 5: Normalized throughput $\mathcal{T}$ across random mixes of $\mathbf{3}$, $\mathbf{4}$, and $\mathbf{5}$ concurrent DNNs.