Sandia National Labs Academic Programs Collaboration Report

Previous Sandia-Illinois collaboration still impacting machine learning field Discovery research can lead to amazing impacts, but the magnitude of those impacts often build over time. A Sandia LDRD Grand Challenge, “Hardware Acceleration of Adaptive Neural Algorithms,” that started in 2015 led to significant contributions in the field of machine learning and 10 associated patents in 2021 alone. Illinois Electrical and Computer Engineering faculty collaborator Steven Levinson, now Professor Emeritus, worked with Sandia PI Conrad James to develop neural theory around symbolic processing, a process that has shed light on data analysis for cybersecurity applications. The new system the team developed integrated neural-inspired machine-learning algorithms with neural-inspired architecture cores, in conventional technology, allowing it to comb through large volumes of data and identify specific data patterns focused precisely on the activity of interest. The initial success of the endeavor, a Temporal Processing Unit, provided the nation with a mechanism for rapidly transforming raw cyber data into information for real time, adaptive recognition and response, but the new technology also provided a neural network categorizer that spun off numerous other deliverables. Since 2015, the HAANA Grand Challenge has resulted in 19 worldwide patents, with one in 2021 belonging to Sandia researchers, Matt Marinella and Sapan Agarwal. Agarwal, who received his undergraduate degree from Illinois and his doctorate from UC Berkeley, originally hired in as a postdoctoral researcher specifically to work on the HAANA Grand Challenge. The HAANA Grand Challenge has resulted in many academic collaborations and workshops, and 39 publications directly associated with the effort.

CONTRIBUTOR Spotlight

Sapan Agarwal, Sandia computer scientist, said of his experience on the HAANA GC, “Through this project, I was able to learn from neural algorithm and architecture experts and became an expert in multi-scale codesign of computing systems from material to algorithms. This [project] has provided me the opportunity to subsequently lead projects in the co-design of heterogenous radiation hardened systems, analog neural accelerators, solvers for linear systems and more.”

57

2021-2022 Collaboration Report