An Adaptive Synaptic Array using Fowler-Nordheim Dynamic Analog Memory
This addresses the energy-efficiency imbalance between training and inference phases in AI systems, representing a domain-specific incremental improvement.
The paper tackles the energy inefficiency in training machine learning systems by introducing a synaptic array using Fowler-Nordheim dynamic analog memory, achieving energy dissipation as low as 5 fJ per update and up to 14-bit programming resolution.
In this paper we present a synaptic array that uses dynamical states to implement an analog memory for energy-efficient training of machine learning (ML) systems. Each of the analog memory elements is a micro-dynamical system that is driven by the physics of Fowler-Nordheim (FN) quantum tunneling, whereas the system level learning modulates the state trajectory of the memory ensembles towards the optimal solution. We show that the extrinsic energy required for modulation can be matched to the dynamics of learning and weight decay leading to a significant reduction in the energy-dissipated during ML training. With the energy-dissipation as low as 5 fJ per memory update and a programming resolution up to 14 bits, the proposed synapse array could be used to address the energy-efficiency imbalance between the training and the inference phases observed in artificial intelligence (AI) systems.