In-Memory Computing Based Hardware Accelerator Module for Deep Neural Networks

被引:0
作者
Appukuttan, Allen [1 ]
Thomas, Emmanuel [1 ]
Nair, Harinandan R. [1 ]
Hemanth, S. [1 ]
Dhanaraj, K. J. [1 ]
Azeez, Maleeha Abdul [1 ]
机构
[1] Natl Inst Technol Calicut, Dept Elect & Commun Engn, Calicut, Kerala, India
来源
2022 IEEE 19TH INDIA COUNCIL INTERNATIONAL CONFERENCE, INDICON | 2022年
关键词
In-Memory Computing; Matrix multiplication; Neural networks; Static Random Access Memory; Von Neumann bottleneck;
D O I
10.1109/INDICON56171.2022.10040126
中图分类号
TP39 [计算机的应用];
学科分类号
081203 ; 0835 ;
摘要
In recent years, AI/ML has been increasingly becoming a part of our daily lives and in the technology around us. With this increasing prevalence, they currently provide the most effective solutions to a wide range of image recognition, speech recognition, and natural language processing challenges. The demand for fast and efficient calculations of increasingly complex models has gained immense importance. One of the biggest technical limitations faced today is the Von Neumann bottleneck. However, the data movement between memory tiers can be reduced by incorporating in-memory computing modules into the architecture. The possibility of using specialized circuits to improve the efficiency of matrix multiplication was attempted on a small test neural network. We focus on using SRAM-based in-memory computing modules, that use binary weights and inputs which can offer a significant improvement to performance and efficiency without a heavy impact on the accuracy of the models.
引用
收藏
页数:6
相关论文
共 13 条
[1]  
Brocard G, 2016, LTSPICE NOUVELLES CO
[2]  
Courbariaux M., 2016, BINARIZED NEURAL NET
[3]  
Davis M, 2021, SNAKEVIZ
[4]  
Anderson AG, 2017, Arxiv, DOI arXiv:1705.07199
[5]  
Jiang HW, 2020, MIDWEST SYMP CIRCUIT, P257, DOI [10.1109/MWSCAS48704.2020.9184455, 10.1109/mwscas48704.2020.9184455]
[6]  
Kim Y, 2021, ELECTRONICS, V10
[7]   Gradient-based learning applied to document recognition [J].
Lecun, Y ;
Bottou, L ;
Bengio, Y ;
Haffner, P .
PROCEEDINGS OF THE IEEE, 1998, 86 (11) :2278-2324
[8]  
Python Documentation, 2021, PYTH DOC PYTH PROF P
[9]   XNOR-Net: ImageNet Classification Using Binary Convolutional Neural Networks [J].
Rastegari, Mohammad ;
Ordonez, Vicente ;
Redmon, Joseph ;
Farhadi, Ali .
COMPUTER VISION - ECCV 2016, PT IV, 2016, 9908 :525-542
[10]  
VONNEUMANN J, 1993, IEEE ANN HIST COMPUT, V15, P28