发明公开
EP4354347A1 SYSTEM AND METHOD FOR ADDITION AND SUBTRACTION IN MEMRISTOR-BASED IN-MEMORY COMPUTING
审中-实审
基本信息:
- 专利标题： SYSTEM AND METHOD FOR ADDITION AND SUBTRACTION IN MEMRISTOR-BASED IN-MEMORY COMPUTING
- 申请号：EP23202358.0 申请日：2023-10-09
- 公开(公告)号：EP4354347A1 公开(公告)日：2024-04-17
- 发明人： REN, Yuan , Wong, Ngai , LI, Can , WANG, Zhongrui
- 申请人： The University of Hong Kong
- 申请人地址： CN Hong Kong Pokfulam Road
- 专利权人： The University of Hong Kong
- 当前专利权人： The University of Hong Kong
- 当前专利权人地址： CN Hong Kong Pokfulam Road
- 代理机构： HGF
- 优先权： US 202263415147 P 2022.10.11
- 主分类号： G06N3/0464
- IPC分类号： G06N3/0464 ; G06N3/065 ; G06N3/048
摘要：
A method of measuring cross-correlation or similarity between input features and filters of neural networks using an RRAM-crossbar architecture to carry out addition/subtraction-based neural networks for in-memory computing. The correlation calculations use L1 norm operations of AdderNet. The RCM structure of the RRAM-cross bar has storage and computing collocated, such that processing is done in the analog domain with low power, low latency and small area. In addition, the impact due to the nonidealities of RRAM device can be alleviated by the implicit ratio-based feature of the structure.
IPC结构图谱:
| G | 物理 |
| --G06 | 计算；推算；计数 |
| ----G06N | 基于特定计算模型的计算机系统 |
| ------G06N3/00 | 基于生物学模型的计算机系统 |
| --------G06N3/02 | .采用神经网络模型 |
| ----------G06N3/04 | ..体系结构，例如，互连拓扑 |
| ------------G06N3/0464 | ...卷积网络 |