Neural-like computing with populations of superparamagnetic basis functions

• DOI: 10.1038/s41467-018-03963-w
• Publication Date: 2018-04-12T11:52:05Z
• Authors: Alice Mizrahi; Tifenn Hirtzlin; Akio Fukushima; Hitoshi Kubota; Shinji Yuasa; Julie Grollier; Damien Querlioz
• Keywords: Nanoelectronics; Neural coding; Basis (linear algebra)

In neuroscience, population coding theory demonstrates that neural assemblies can achieve fault-tolerant information processing. Mapped to nanoelectronics, this strategy could allow for reliable computing with scaled-down, noisy, imperfect devices. Doing so requires the components form a set of basis functions in terms their response inputs, offering physical substrate computing. Such be implemented CMOS technology, but corresponding circuits have high area or energy requirements. Here, we show nanoscale magnetic tunnel junctions instead assembled meet these We demonstrate experimentally nine implement functions, providing data achieve, example, generation cursive letters. design hybrid magnetic-CMOS systems based on interlinked populations and they learn realize non-linear variability-resilient transformations low imprint power.