Materials for neuromorphic computing - enhancing spin orbit torque efficiency through low energy mixed ion bombardment.
The rapid advancement towards Artificial Intelligence through the 21st century begs the question of whether our current computing facility is able to fully meet the computing demands of the future. Indeed, as many have pointed out, the current computing architecture is soon to plateau, attribu...
Saved in:
Main Author: | |
---|---|
Other Authors: | |
Format: | Final Year Project |
Language: | English |
Published: |
Nanyang Technological University
2023
|
Subjects: | |
Online Access: | https://hdl.handle.net/10356/172131 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Nanyang Technological University |
Language: | English |
Summary: | The rapid advancement towards Artificial Intelligence through the 21st century begs
the question of whether our current computing facility is able to fully meet the
computing demands of the future. Indeed, as many have pointed out, the current
computing architecture is soon to plateau, attributed to the inherent throughput
limitations of the von Neumann architecture. In this thesis, we discuss the
Neuromorphic computing paradigm – a brain-inspired computing framework that
circumvents the von Neumann bottleneck through co-locating the memory and
computing subsystems. We illustrate how non-volatile memory based on spintronics
has potential applications (beyond consumer electronics) towards realising non volatile magnetic memory using Spin Orbit Torque (SOT). Here, we also delve into
the physics of magnetic materials which facilitate SOT. This gives us the grounding to
understand our investigation to enhance SOT using a novel method of low-energy
mixed ion exposure. Our results with Pt/Co/W structures have shown significant
positive enhancement toward SOT efficiency, which provides future researchers with
an extra parameter to finely tune SOT in future work |
---|