Neuromorphic computing promises to dramatically improve the efficiency of certain computational tasks, such as perception and decision making. While software and specialized hardware implementations of neural networks have made tremendous progress, both implementations are still many orders of magnitude less energy efficient than the human brain. This talk will introduce the current state of neuromorphic computing and discuss the recent hardware advances that have bolstered the progress of machine learning. I will discuss how a new superconducting platform, based on dynamically reconfigurable magnetic Josephson junctions with spiking energies less than one attojoule, could lead to large-scale neuromorphic systems with better energy efficiency than the human brain.
NSCI Committee
1:00 p.m. - 2:00 p.m. (Gaithersburg, Bldg. 221, Room B145)
11:00 a.m. - 12:00 p.m. (Boulder, VTC in 81-1A116)
Michael Schneider, NIST/Boulder
Outside attendees need to contact Barry Schneider in order to obtain the site badges required to enter NIST grounds and to attend the seminar. 24 hour notice is required for US citizens and 3 days for non-US citizens. Please contact bis [at] nist.gov (bis[at]nist[dot]gov) to be added to the visitor list. Visitors must check in at the NIST Visitor Center to pick up their badges. A photo ID is required for US citizens and a passport or green card for foreign nationals. There is also the possibility of viewing the seminar as a webcast. Again, please contact Barry Schneider for details preferably by email.