Swan: A Neural Engine for Efficient DNN Training on Smartphone SoCs
This addresses the need for efficient on-device training to enhance data privacy and reduce communication overheads for smartphone users, representing a domain-specific incremental improvement.
The paper tackles the problem of inefficient DNN training on smartphone SoCs, which can degrade user experience, and presents Swan, a neural engine that improves performance by 1.2 to 23.3 times over state-of-the-art methods.
The need to train DNN models on end-user devices (e.g., smartphones) is increasing with the need to improve data privacy and reduce communication overheads. Unlike datacenter servers with powerful CPUs and GPUs, modern smartphones consist of a diverse collection of specialized cores following a system-on-a-chip (SoC) architecture that together perform a variety of tasks. We observe that training DNNs on a smartphone SoC without carefully considering its resource constraints can not only lead to suboptimal training performance but significantly affect user experience as well. In this paper, we present Swan, a neural engine to optimize DNN training on smartphone SoCs without hurting user experience. Extensive large-scale evaluations show that Swan can improve performance by 1.2 - 23.3x over the state-of-the-art.