Digital Biologically Plausible Implementation of Binarized Neural Networks with Differential Hafnium Oxide Resistive Memory Arrays.

Brains perform intelligent tasks with extremely low energy consumption, probably to a large extent because they merge computation and memory entirely, and because they function using low precision computation. The emergence of resistive memory technologies provides an opportunity to integrate logic and memory tightly in hardware. In parallel, the recently proposed concept of Binarized Neural Network, where multiplications are replaced by exclusive NOR logic gates, shows a way to implement artificial intelligence using very low precision computation. In this work, we therefore propose a strategy to implement low energy Binarized Neural Networks, which employs these two ideas, while retaining energy benefits from digital electronics. We design, fabricate and test a memory array, including periphery and sensing circuits, optimized for this in-memory computing scheme. Our circuit employs hafnium oxide resistive memory integrated in the back end of line of a 130 nanometer CMOS process, in a two transistors - two resistors cell, which allows performing the exclusive NOR operations of the neural network directly within the sense amplifiers. We show, based on extensive electrical measurements, that our design allows reducing the amount of bit errors on the synaptic weights, without the use of formal error correcting codes. We design a whole system using this memory array. We show on standard machine learning tasks (MNIST, CIFAR-10, ImageNet and an ECG task) that the system has an inherent resilience to bit errors. We evidence that its energy consumption is attractive with regards to more standard approaches, and that it can use the memory devices in regimes where they exhibit particularly low programming energy and high endurance. We conclude the work by discussing its associations between biological plausible ideas and more traditional digital electronics concepts.