Design of RRAM-Based Integrate and Fire Neuron And Programmable Synapse for Neuromorphic Computing
No Thumbnail Available
Date
2024
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
A human brain can perform compute-intensive tasks, such as multi-object recognition, reasoning, and decision-making, consuming only 20 W power. Whereas, to recognize 1000 different objects, a CPU consumes around 250 W power. Around 1011 neurons in the human brain are interconnected through approximately 1015 synapses responsible for the brain’s exceptional computing capacity. The advancements in processing technology have reduced the technology nodes drastically, which further reduced the power consumption of the processors; still, they cannot match the low power consumption of the human brain. Even with the latest technological advancements, optimizing the processors with Von Neumann architectures for speed and power becomes challenging because of the memory Bottleneck
Description
Supervisor: Trivedi, Gaurav
Keywords
Neuromorphic Computing, RRAM, Integrate and Fire Neuron, Synapse