A 65nm 1Mb nonvolatile computing-in-memory ReRAM macro with sub-16ns multiply-and-accumulate for binary DNN AI edge processors

By Wei-Hao Chen, Kai-Xiang Li, Wei-Yu Lin, Kuo-Hsiang Hsu, Pin-Yi Li, Cheng-Han Yang, Cheng-Xin Xue, En-Yu Yang, Yen-Kai Chen, Yun-Sheng Chang and 7 more. IEEE International Solid-State Circuits Conference, 2018

300 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Electrical and Electronic Engineering

View Paper: https://www.semanticscholar.org/paper/e6fbb8d3bededb92d1e311fbb6b2625453fc6642

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

1. Huaqiang Wu
2. Naveen Verma
3. Bin Gao
4. Wenqiang Zhang
5. Peng Yao
6. Jianshi Tang
7. Qing Wu
8. He Qian

In Electrical and Electronic Engineering

1. Shimeng Yu
2. J. Joshua Yang
3. Qiangfei Xia
4. Huaqiang Wu
5. Zhongrui Wang
6. Bin Gao
7. He Qian
8. Jianshi Tang

Citations in the corpus

• Hardware/Software Co-Exploration of Neural Architectures (2019, unknown): “Second, innovations on computing architecture achieves great success for executing inference phase of neural networks [44], we plan to apply the proposed framework to co-explore neural architectures with the novel computing architectures (e.g., computing-in-memory).”
• Three-dimensional memristor circuits as complex neural networks (2020, methodology): “It has been implemented in standard 2D memristor crossbar arrays, but each array can only process convolutions from the same input pixels due to the aforementioned full connectivity 31,32 .”
• In-Memory Computing: Advances and prospects (2019, background): “Thus, IMC can apply to any memory technology integrated in such a structure [8]–[10].”
• Neuro-inspired computing chips (2020, methodology): “Although applying heuristic compensation by adding scaling factors to the output terminal can mitigate this effect for some specific neural networks 66 , The past milestones 11,12,17,21,23,63,71,105 and future prospects are illustrated.”