Jeremy Love, "Low Power Intelligent Algorithms for Detector Front-Ends"

Tuesday, February 11, 2020 from to (US/Central)
at Building 362 ( F-108 )
Abstract:  This seminar will present an Early Career Research proposal for the work necessary to make low power intelligent detector front-ends possible.  Neuromorphic accelerators offer the best possibility to provide intelligence in the high radiation power constrained environments of detector front-ends electronics.  The devices are designed for sparsely encoded spiking neural networks.  This proposal is for the work necessary to extending existing software frameworks for hyperparameter scans that include power usage.  It will be used to develop and test algorithms for low power self-calibrating and intelligent detectors and verify their performance in real world applications.
Go to day