Using Loihi for Neuromorphic Education

Panel

Alex Dimitrov, WSU

Terry Stewart, National Research Council of Canada

Johan Kwisthout, Radboud University

Konstantinos Michmizos, Rutgers University

Facilitators:

@David Florey (Deactivated) @Jon Tse (Unlicensed) @Sumedh Risbud

Session Structure

Panel speakers to each spend 5-10 minutes presenting their experiences teaching neuromorphic computing and how they’ve incorporated software/hardware solutions into their curriculum. We will then discuss what has worked well, where additional support (materials, software, hardware) could be helpful and potential for moving toward a repository of material others could draw from/contribute to.

Speaker Bios

@alex.dimitrov

Alex Dimitrov is associate professor of mathematics and neuroscience at Washington State University whose research interests include the study of neural information processing, neural coding and information representation in biological systems.

@Johan Kwisthout

Johan Kwisthout is associate professor and AI degree programme director at the Donders Institute at Radboud University. He teaches a graduate level course on neuromorphic computing since 2019. His research interests are mostly in the formal assessment of the computational power of neuromorphic architectures, with 'energy' as a resource on par with the more traditional 'time' and 'space' known from classical computational complexity theory.

@Konstantinos

Konstantinos Michmizos is an Assistant Professor in the Department of Computer Science at Rutgers University, where he directs the Computational Brain Lab. He joined Rutgers in 2015, upon completing two postdoctoral trainings on robotics, at MIT, and computational neuroscience, at Harvard Medical School. His research on neuromorphic computing and robotics is supported by Intel and NIH, among other agencies. He has been teaching a brain-inspired computing course for 5 years including term projects using Loihi.

@Terry Stewart

Terry Stewart is a Research Officer with the National Research Council of Canada, where he develops large-scale computational neuroscience models, and applies them to industry-relevant tasks. He was a primary developer on Nengo (a software stack supporting a variety of neuromorphic hardware), worked on the initial demos for Intel's digital neuromorphic chip Loihi, and helped create Stanford's analog neuromorphic chip Braindrop.

Audience

Please be ready with your questions, comments and experiences you would like to share.