Lecture
Hardware-Software Co-Design towards Efficient Neuromorphic Computing
- Qinyu Chen
- Date
- Tuesday 14 February 2023
- Time
- Address
-
Snellius
Niels Bohrweg 1
2333 CA Leiden - Room
- 407-409
Artificial Intelligence (AI) has been making significant progress in recent years and reshaping modern society. Deep learning and neuromorphic computing are two major paths in the development of AI, with deep learning being more focused on computer science and mathematics, and neuromorphic computing taking inspiration from the structure and function of the human brain. Spiking neural networks (SNNs) are a type of neural network that falls under the umbrella of neuromorphic computing, and they have shown great potential for energy efficiency compared to traditional deep neural networks (DNNs) due to their biologically inspired computation and are expected to be part of the future AI portfolio. However, silicon implementations of SNNs have generally lagged behind state-of-the-art silicon implementations of DNNs. This talk will focus on the topic of efficient SNN architecture through hardware-software co-design, which is an important aspect of the development of neuromorphic computing and the future of AI.