RNSnet: In-Memory Neural Network Acceleration Using Residue Number System - Sahand Salamat - ICRC 2018

337 views

IEEE Future Directions

Create Account or Sign In to post comments

#IEEE #ICRC #Rebooting Computing #2018 #conference #event #computing #technology #neural networks #applications #processing #memory #computation #Residue Number System #RNS #RNS net #digital computation

Sahand Salamat, from the University of California San Diego and System Energy Efficiency Lab, looks at deep neural networks and their role in the majority of applications in use today. Salamat points to the main issue of overloaded processing and presents solutions using RNS, explaining the technical details behind these memory computations.

Sahand Salamat, from the University of California San Diego and System Energy Efficiency Lab, looks at deep neural networks and their role in the majority of applications in use today. Salamat points to the main issue of overloaded processing and presents solutions using RNS, explaining the technical details behind these memory computations.

March 4, 2019

Next Up

00:30:25

2023 Future of Telepresence Workshop: Measuring Latency-Induced Cognitive Load for Teleoperated Rescue - Dr. John Blitch

00:27:16

2023 Future of Telepresence Workshop: Teleoperation, Telepresence, and Autonomy: Synergy and Complementarity - Paolo Fiorini

00:32:41

IEEE LEO SatS: Tbit/s Satellite Communication Links Based on Optical Technologies - Yannik Horst

00:31:19

IEEE LEO SatS: Inter-Satellite Data Relay Service (IDRS) - Eyal Trachtman

00:30:11

IEEE LEO SatS: Legacy Ground Stations and Ground Operations - Marcin Gnat

00:10:59

IEEE LEO SatS: Fall 2023 Workshop on LEO Ground Stations & Ground Station Services Intro - Jan Budroweit