Printed Stochastic Computing Neural Networks
Dennis D. Weller1, Nathaniel Bleier2, Michael Hefenbrock1, Jasmin Aghassi-Hagmann3, Michael Beigl1, Rakesh Kumar2 and Mehdi B. Tahoori1
1Karlsruhe Institute of Technology
2University of Illinois Urbana-Champaign
3Offenburg University of Applied Sciences
ABSTRACT
Printed electronics (PE) offers flexible, extremely lowcost, and on-demand hardware due to its additive manufacturing process, enabling emerging ultra-low-cost applications, including machine learning applications. However, large feature sizes in PE limit the complexity of a machine learning classifier (e.g., a neural network (NN)) in PE. Stochastic computing Neural Networks (SCNNs) can reduce area in silicon technologies, but still require complex designs due to unique implementation tradeoffs in PE. In this paper, we propose a printed mixed-signal system, which substitutes complex and power-hungry conventional stochastic computing (SC) components by printed analog designs. The printed mixed-signal SC consumes only 35% of power consumption and requires only 25% of area compared to a conventional 4-bit NN implementation. We also show that the proposed mixedsignal SC-NN provides good accuracy for popular neural network classification problems. We consider this work as an important step towards the realization of printed SC-NN hardware for nearsensor- processing.
Keywords: Printed Electronics, Stochastic Computing, Neural Networks, Electrolyte-Gated Transistors.
