Georgia Institute of Technology
Adaptive Oxide Based Low-Power Memristive Devices for Neuromorphic Computing
Abstract
dc:description.abstractThis thesis aims to develop a hardware platform using intercalation based lithium niobite memristors to implement scalable neuromorphic architectures with high energy-efficiency and co-localized processing and main memory, thus overcoming the memory wall. Brain-inspired neuromorphic computing is a crucial field in addressing the increased need for collection, analysis and decision making from high volumes of dynamic unstructured data generated globally, at low power consumption. Memristive devices have emerged a key enabling technology for developing such large scale neuromorphic computing platforms. Lithium niobite is an adaptive suboxide which has shown promise in developing volatile and non-volatile memristors for highly scalable and low-power neuromorphic circuitry. The culmination of this work has demonstrated a memristive technology that implements all major functions of a neural network: analog training, linear resistive changes, temporal tuning, varied temporal response and adaptive activation, with each efficiently implemented in hardware.
Degree
thesis:*- Level thesis:degree_level
- Doctoral
- Department dc:contributor.department
- Electrical and Computer Engineering
- Grantor dc:publisher
- Georgia Institute of Technology
- Year dc:date.issued
- 2023
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Ghosh, Aheli
- Advisor dc:contributor.advisor
-
- Doolittle, William Alan
- Committee members dc:contributor.committeemember
-
- Khan, Asif I.
- Hunt, William D.
- Yoder, P. Doug
- Vogel, Eric M.
Subjects
dc:subject × 5Rights
- Language dc:language.iso
- en_US
Identifiers
dc:identifier.*- Handle dc:identifier.uri
- https://hdl.handle.net/1853/72702