A 28nm 16.9-300TOPS/W Computing-in-Memory Processor Supporting Floating-Point NN Inference/Training with Intensive-CIM Sparse-Digital Architecture

📋 论文概要

该论文提出了一款28nm工艺的存算一体处理器，支持浮点神经网络推理与训练，能效达到16.9-300TOPS/W。针对传统浮点CIM需要复杂逻辑或长对齐周期的问题，提出了高效浮点运算方案。

💡 主要创新点

• 提出了一种支持浮点乘累加运算的存算一体架构，避免了传统方案中的长对齐延迟。
• 实现了16.9-300TOPS/W的宽能效范围，兼顾高性能与低功耗。
• 可能采用了Intensive-CIM技术，优化浮点数据在存储阵列中的映射与计算。

🏷 关键词

📄 原文摘要

University, Beijing, China 1 2 Computing-in-memory (CIM) has shown high energy efficiency on low-precision integer multiply-accumulate (MAC) [1-3]. However, implementing floating-point (FP) operations using CIM has not been thoroughly explored. Previous FP CIM chips [4-5] require either complex in-memory FP logic or have lengthy alignment-cycle latencies arising from converting FP data having different exponents into integer data. The challenges for an energy-efficient and accurate FP CIM processor are shown in Fig. 16.3.1. Firstly, aligning an FP vector onto a CIM module requires a long bit-serial sequence due to infrequent but long tail values, incurring many CIM cycles. In this work, we observe that most exponents of FP data are clustered in a small range, which motivates dividing FP

👥 作者与机构

Jinshan Yue1, Chaojie He1, Zi Wang1, Zhaori Cong1, Yifan He2, Mufeng Zhou2,

Wenyu Sun2, Xueqing Li2, Chunmeng Dou1, Feng Zhang1, Huazhong Yang2, Yongpan Liu2, Ming Liu1 Institute of Microelectronics of the Chinese Academy of Sciences, Beijing, China Tsinghua