COGNITIVE COMPUTING WITH NANOSCALE ELECTRONIC SYNAPSES: MERGING NANOELECTRONICS AND NEUROSCIENCE
Due to physical limits, CMOS transistor scaling has started to face significant barriers to achieving historical performance improvements. Addition-ally, the conventional computing paradigm based on Von Neumann architecture becomes increasingly inefficient as the computation gets more complex. In contrast, the human brain is massively parallel, robust, low power and fault-tolerant, hence much more efficient in complex operations compared to conventional hardware. Recent findings from biological experiments as well as computational studies suggest that learning behavior in the human brain might be a result of changes in synaptic strength. This suggests that an electronic device that emulates synapses inside the brain will be the most important element of a human-scale intelligent system. Due to their gradual resistance abilities, PCM (phase change memory) and RRAM (resistive random access memory) have been demonstrated to emulate the learning behavior of biological synapses. Mr. Eryilmaz will explain how these devices have the potential to open up the last frontier in computing, human-scale intelligent systems, which has not yet been realized due to the lack of efficient hardware.
Speaker: Sukru Burc Eryilmaz, Stanford
Thursday, 04/18/13
Contact:
Website: Click to VisitCost:
FreeSave this Event:
iCalendarGoogle Calendar
Yahoo! Calendar
Windows Live Calendar
Lockheed Martin Colloquia
Building 202 Auditorium
Palo Alto, CA 94304
Website: Click to Visit
