GreenAuto: An Automated Platform for Sustainable AI Model Design on Edge Devices

TL;DR

GreenAuto is an automated platform that efficiently explores and identifies energy-efficient AI models for edge devices using a Pareto front search guided by energy predictors, promoting sustainability.

Contribution

It introduces a fully automated, energy-aware neural architecture search platform that optimizes AI models for sustainability on edge devices.

Findings

Successfully identifies sustainable models with minimal human intervention

Uses energy predictors to guide the search process effectively

Demonstrates improved energy efficiency in AI models for edge deployment

Abstract

We present GreenAuto, an end-to-end automated platform designed for sustainable AI model exploration, generation, deployment, and evaluation. GreenAuto employs a Pareto front-based search method within an expanded neural architecture search (NAS) space, guided by gradient descent to optimize model exploration. Pre-trained kernel-level energy predictors estimate energy consumption across all models, providing a global view that directs the search toward more sustainable solutions. By automating performance measurements and iteratively refining the search process, GreenAuto demonstrates the efficient identification of sustainable AI models without the need for human intervention.