Affiliation:
1. Georgia Institute of Technology, Atlanta, United States
2. Lawrence Livermore National Laboratory, Livermore, United States
Abstract
The prediction accuracy of deep neural networks (DNNs) deployed at the edge can deteriorate over time due to shifts in the data distribution. For heightened robustness, it’s crucial for DNNs to continually refine and improve their predictive capabilities. However, adaptation in resource-limited edge environments is fraught with challenges: (i) new labeled data might be unavailable; (ii) on-device adaptation is a necessity as cloud connections may be inaccessible; and (iii) the adaptation procedure should prioritize speed, memory efficiency, and energy conservation. Compute-In-Memory (CIM) has recently garnered attention for its computational efficacy and superior operational bandwidth. Additionally, emerging lightweight unsupervised DNN adaptation techniques during test-time have showcased promising results in enhancing model accuracy for data with noise. This article pioneers a holistic benchmarking exploration of these methods, assessing their performance and energy efficacy across diverse CIM architectures in edge and autonomous systems. Our findings reveal that the proposed adaptation strategies can adapt to both environment shifts and inherent hardware noise. Engaging in a thorough cross-layer algorithm-hardware-technology co-design space exploration, we highlight pivotal trade-offs among accuracy, performance, and energy for various DNN adaptation techniques and CIM configurations.
Funder
CoCoSys
Semiconductor Research Corporation
Defense Advanced Research Projects Agency
Publisher
Association for Computing Machinery (ACM)
Reference75 articles.
1. CiMLoop: A flexible, accurate, and fast compute-in-memory modeling tool;Andrulis Tanner;arXiv preprint arXiv:2405.07259,2024
2. Combined HW/SW Drift and Variability Mitigation for PCM-Based Analog In-Memory Computing for Neural Network Applications
3. Autonomous Navigation via Deep Reinforcement Learning for Resource Constraint Edge Nodes Using Transfer Learning
4. Hamza Errahmouni Barkam, Sanggeon Yun, Paul R. Genssler, Zhuowen Zou, Che-Kai Liu, Hussam Amrouch, and Mohsen Imani. 2023. HDGIM: Hyperdimensional genome sequence matching on unreliable highly scaled FeFET. In 2023 Design, Automation & Test in Europe Conference & Exhibition (DATE). IEEE, 1–6.
5. A Multilayer Neural Accelerator With Binary Activations Based on Phase-Change Memory