A new technical paper titled “Self-organization of an inhomogeneous memristive hardware for sequence learning” was just published by researchers at University of Zurich, ETH Zurich, Université Grenoble Alpes, CEA, Leti and Toshiba.
“We design and experimentally demonstrate an adaptive hardware architecture Memristive Self-organizing Spiking Recurrent Neural Network (MEMSORN). MEMSORN incorporates resistive memory (RRAM) in its synapses and neurons which configure their state based on Hebbian and Homeostatic plasticity respectively,” states the paper’s abstract.
Find the technical paper here. Published October 2022.
Payvand, M., Moro, F., Nomura, K. et al. Self-organization of an inhomogeneous memristive hardware for sequence learning. Nat Commun 13, 5793 (2022). https://doi.org/10.1038/s41467-022-33476-6.
Related Reading
Spiking Neural Network (SNN) Knowledge Center
Neuromorphic Computing: Challenges, Opportunities Including Materials, Algorithms, Devices & Ethics
MEMprop: Gradient-Based Learning To Train Fully Memristive SNNs
11 Ways To Reduce AI Energy Consumption
Pushing AI to the edge requires new architectures, tools, and approaches.
Leave a Reply