Scheduling Real-time Deep Learning Services as Imprecise Computations

TL;DR

This paper introduces a real-time scheduling algorithm for deep learning tasks on edge devices, optimizing accuracy by selectively executing optional neural network parts within deadline constraints.

Contribution

It proposes a novel scheduling approach that treats neural network workflows as imprecise computations with mandatory and optional parts, improving accuracy under real-time constraints.

Findings

Increases accuracy by 10-20%

Nearly eliminates deadline misses

Effective on GPU hardware for vision tasks

Abstract

The paper presents an efficient real-time scheduling algorithm for intelligent real-time edge services, defined as those that perform machine intelligence tasks, such as voice recognition, LIDAR processing, or machine vision, on behalf of local embedded devices that are themselves unable to support extensive computations. The work contributes to a recent direction in real-time computing that develops scheduling algorithms for machine intelligence tasks with anytime prediction. We show that deep neural network workflows can be cast as imprecise computations, each with a mandatory part and (several) optional parts whose execution utility depends on input data. The goal of the real-time scheduler is to maximize the average accuracy of deep neural network outputs while meeting task deadlines, thanks to opportunistic shedding of the least necessary optional parts. The work is motivated by the proliferation of increasingly ubiquitous but resource-constrained embedded devices (for applications ranging from autonomous cars to the Internet of Things) and the desire to develop services that endow them with intelligence. Experiments on recent GPU hardware and a state of the art deep neural network for machine vision illustrate that our scheme can increase the overall accuracy by 10%-20% while incurring (nearly) no deadline misses.