Room 830, Building J03, Maze Cresent
Tel: +61 2 9351 2014 | Fax: +61 2 9351 3847
Jason Ang received his B.E. (Hons) in Electrical Engineering from the University of Melbourne in 1998.
He was with Advanced Micro Devices (AMD) working on non-volatile memory devices from 1999-2000. From 2000-2007 he was with Agilent/Avago Technologies working on optical and laser mouse sensors integrated circuits design.
He is currently a research student at the University of Sydney, working on neural network-based memory architecture.
The research work currently involved in is to develop a neural network-based memory architecture targeting an FPGA platform for application in the area of pattern recognition.
The architecture is based on a spiking neural network model, the third generation of neural network models which incorporate the concept of time in addition to the neuronal and synaptic states simulated in previous generations of neural networks.
The novel aspect of this work is the exploitation of the massive interconnect resources in an FPGA as programmable delay lines mimicking the axonal connections in biological neural networks. Through delay adaptation, the system is expected to train itself for storing patterns in the form of spatiotemporal spike sequences and retrieve them back when prompted.
- Digital systems design
- Design with Hardware Description Language (HDL)
- Hardware/software co-design
- Reconfigurable computing