Multi-Objective Neural Architecture Search for In-Memory Computing
This work addresses efficiency challenges for deploying ML models on in-memory computing systems, but it is incremental as it builds on existing NAS and CNN methods.
The paper tackled the problem of deploying machine learning tasks efficiently on in-memory computing architectures by using neural architecture search to optimize convolutional neural networks, achieving high accuracy with reduced latency and energy consumption across three image classification datasets.
In this work, we employ neural architecture search (NAS) to enhance the efficiency of deploying diverse machine learning (ML) tasks on in-memory computing (IMC) architectures. Initially, we design three fundamental components inspired by the convolutional layers found in VGG and ResNet models. Subsequently, we utilize Bayesian optimization to construct a convolutional neural network (CNN) model with adaptable depths, employing these components. Through the Bayesian search algorithm, we explore a vast search space comprising over 640 million network configurations to identify the optimal solution, considering various multi-objective cost functions like accuracy/latency and accuracy/energy. Our evaluation of this NAS approach for IMC architecture deployment spans three distinct image classification datasets, demonstrating the effectiveness of our method in achieving a balanced solution characterized by high accuracy and reduced latency and energy consumption.