To achieve this, our researchers develop AI hardware solutions that capitalise on a new electronic fabric – memristors - that offers more functionality for less energy and chip area.
Our Hardware systems for AI deliver state-of-the art capability in:
- real-time/on-node data sensing/processing, in response to the power/bandwidth constraints of “continuously on” systems;
- efficient in-situ pattern recognition systems, to enhance computation capability locally;
- efficientmemorymodules,formitigatingfinitememoryresourcesconstraintsoflife-longlearning/continuously- on systems;
- systems capable of reasoning, for consolidating distinct and often dissimilar features/inputs of hardware systems;
You can also watch this video on YouTube