Brain-inspired technologies for efficient AI: From silicon eyes to neuromorphic computers
Date:
In order to process numbers and make decisions very quickly, current AI systems require powerful computers and enormous energy resources. On the other hand, although our brain cannot perform 1,000,000 multiplications in less than a second, it is capable of making decisions in real-life scenarios that are far more robust than any man-made system. Brain-inspired technologies such as neuromorphic systems developed by humans are getting closer and closer to brain-like processing with a fraction of the energy of today’s systems. By utilising principles known to exist in the brain, neuromorphic systems are the new frontier of AI, combining robustness and efficiency at a cost of just a few milliwatts. The lecture will present neuromorphic systems and AI applications developed at OHM.