Link to the University of Pittsburgh Homepage
Link to the University Library System Homepage Link to the Contact Us Form

Memristor-Based Analog Neuromorphic Computing Engine Design and Robust Training Scheme

Liu, Beiye (2014) Memristor-Based Analog Neuromorphic Computing Engine Design and Robust Training Scheme. Master's Thesis, University of Pittsburgh. (Unpublished)

Primary Text

Download (1MB) | Preview


The invention of neuromorphic computing architecture is inspired by the working mechanism of human-brain. Memristor technology revitalized neuromorphic computing system design by efficiently executing the analog Matrix-Vector multiplication on the memristor-based crossbar (MBC) structure. In this work, we propose a memristor crossbar-based embedded platform for neuromorphic computing system. A variety of neural network algorithms with threshold activation function can be easily implemented on our platform. However, programming the MBC to the target state can be very challenging due to the difficulty to real-time monitor the memristor state during the training. In this thesis, we quantitatively analyzed the sensitivity of the MBC programming to the process variations and input signal noise. We then proposed a noise-eliminating training method on top of a new crossbar structure to minimize the noise accumulation during the MBC training and improve the trained system performance, i.e., the pattern recall rate. A digital-assisted initialization step for MBC training is also introduced to reduce the training failure rate as well as the training time. We also proposed a memristor-based bidirectional transmission exhibition/inhibition synapse and implemented neuromorphic computing demonstration with our proposed synapse. Experiment results show that the proposed design has high tolerance on process variation and input noise. Different benefits of MBC system and new synapse-based system will be compared in our thesis.


Social Networking:
Share |


Item Type: University of Pittsburgh ETD
Status: Unpublished
CreatorsEmailPitt UsernameORCID
Liu, Beiyebel34@pitt.eduBEL34
ETD Committee:
TitleMemberEmail AddressPitt UsernameORCID
Committee ChairChen, YIC52
Committee MemberHai, HAL66
Committee MemberMingui, Sundrsun@pitt.eduDRSUN
Committee MemberZhihong, ZHM4
Date: 16 June 2014
Date Type: Publication
Defense Date: 18 March 2014
Approval Date: 16 June 2014
Submission Date: 3 April 2014
Access Restriction: No restriction; Release the ETD for access worldwide immediately.
Number of Pages: 49
Institution: University of Pittsburgh
Schools and Programs: Swanson School of Engineering > Electrical Engineering
Degree: MS - Master of Science
Thesis Type: Master's Thesis
Refereed: Yes
Uncontrolled Keywords: memristor, neuromorphic, low power
Date Deposited: 16 Jun 2014 18:09
Last Modified: 15 Nov 2016 14:18


Monthly Views for the past 3 years

Plum Analytics

Actions (login required)

View Item View Item