Cluster-based Associative Memories Built From Unreliable Storage

François Leduc-Primeau (1) , Vincent Gripon (2, 3) , Michael Rabbat (1) , Warren Gross (1)

1 McGill University = Université McGill [Montréal, Canada]
2 ELEC - Département Electronique
3 Lab-STICC_TB_CACS_IAS

We consider associative memories based on clustered graphs that were recently introduced. These memories are almost optimal in terms of the amount of storage they require (efficiency), and allow retrieving messages with low complexity. We study an unreliable implementation of the memory and compare its error rate and storage efficiency with that of a reliable implementation. We present analytical and simulation results that indicate that the proposed memory structure can tolerate a large number of faults at a reasonable cost, thereby making it a good candidate for achieving highly efficient circuit implementations of associative memories. View full abstract"

Domaines: Neurosciences

François Leduc-Primeau, Vincent Gripon, Michael Rabbat, Warren Gross. Cluster-based Associative Memories Built From Unreliable Storage. ICASSP 2014 : IEEE International Conferences on Acoustics, Speech and Signal Processing, May 2014, Florence, Italy. pp.8370-8374, ⟨10.1109/ICASSP.2014.6855234⟩. ⟨hal-01057357⟩