RecPIM: Efficient In-Memory Processing for Personalized Recommendation Inference Using Near-Bank Architecture
Sign Up to like & getrecommendations! Published in 2024 at "IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems"
DOI: 10.1109/tcad.2024.3386117
Abstract: Deep learning (DL)-based personalized recommendation systems consume the major resources in modern AI data centers. The embedding layers with large memory capacity requirement and high bandwidth demand have been identified as the bottleneck of personalized… read more here.
Keywords: memory; personalized recommendation; memory processing; recommendation inference ... See more keywords