FAT: an in-memory accelerator with fast addition for ternary weight neural networks

• Main Authors: Zhu, Shien, Duong, Luan H. K., Chen, Hui, Liu, Di, Liu, Weichen
• Other Authors: School of Computer Science and Engineering
• Format: Article
• Language: English
• Published: 2022
• Subjects: Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence; Engineering::Computer science and engineering::Hardware::Arithmetic and logic structures; Engineering::Computer science and engineering::Computer systems organization::Special-purpose and application-based systems; Ternary Wight Neural Network; In-Memory Computing; Convolutional Neural Network; Spin-Transfer Torque Magnetic Random-Access Memory
• Online Access: https://hdl.handle.net/10356/162483
• Institution: Nanyang Technological University

Convolutional Neural Networks (CNNs) demonstrate excellent performance in various applications but have high computational complexity. Quantization is applied to reduce the latency and storage cost of CNNs. Among the quantization methods, Binary and Ternary Weight Networks (BWNs and TWNs) have a uni...