Members will Receive Resources for Exploring Neuromorphic Computing Use Cases
By Dr. Michael Mayberry
This week, we hosted the Neuro Inspired Computational Elements (NICE) workshop at our Oregon campus with the goal of bringing together researchers from different scientific disciplines to discuss and explore the development of next-generation computing architectures, including neuromorphic computing. Today at the workshop, we provided an update on Intel’s neuromorphic research and announced a collaborative research initiative to encourage experimentation with our Loihi neuromorphic test chip.
Here’s a status of our neuromorphic computing efforts and details on this new research community.
Where We Are
Fabrication and packaging of our Loihi test chip was completed in early November, and we began power-on and validation. We were pleased to find 100 percent functionality, a wide operating margin and few bugs overall. Our small-scale demonstrations that we had prepared on our emulator worked as expected on the real silicon, though, of course, running orders of magnitude faster. Our equivalent of a “Hello World” application is recognizing a 3-D object from multiple viewing angles, structured after the COIL-20 example from Columbia University. As measured at our lab, this particular application uses less than 1 percent of Loihi, learns the training set in seconds and consumes tens of milliwatts.
We shared Loihi architectural details in a paper that IEEE Micro recently published, and we presented those details and several demos to NICE workshop attendees this week.
We have delivered the first developer systems to select research collaborators who are working on a variety of applications including sensing, motor control, information processing and more. Software development tools remain one of our focus areas, and we’re looking forward to running much larger scale applications in conjunction with research collaborators. As we learn more together, we expect progress to accelerate, and that’s where today’s announcement comes in.