Achieving ideal accuracies in analog neuromorphic computing using periodic carry

Sapan Agarwal, Robin B. Jacobs Gedrim, Alexander H. Hsia, David Russell Hughart, Elliot J. Fuller, A. Alec Talin, Conrad D. James, Steven J. Plimpton, Matthew Marinella

2017 · 62 citations · 4 references

DOIFull text

Open access

Concepts

Abstract

Analog resistive memories promise to reduce the energy of neural networks by orders of magnitude. However, the write variability and write nonlinearity of current devices prevent neural networks from training to high accuracy. We present a novel periodic carry method that uses a positional number system to overcome this while maintaining the benefit of parallel analog matrix operations. We demonstrate how noisy, nonlinear TaO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">x</sub> devices that could only train to 80% accuracy on MNIST, can now reach 97% accuracy, only 1% away from an ideal numeric accuracy of 98%. On a file type dataset, the TaO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">x</sub> devices achieve ideal numeric accuracy. In addition, low noise, linear Li <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1-x</sub> CoO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> devices train to ideal numeric accuracies using periodic carry on both datasets.

References

4