Search

Abstract Physical reservoir computing is a promising solution for accelerating artificial intelligence (AI) computations. Various physical systems that exhibit nonlinear and fading-memory properties have been proposed as physical reservoirs. Highly-integrable physical reservoirs, particularly for edge AI computing, has a strong demand. However, realizing a practical physical reservoir with high performance and integrability remains challenging. Herein, we present an analogue circuit reservoir with a simple cycle architecture suitable for complementary metal-oxide-semiconductor (CMOS) chip integration. In several benchmarks and demonstrations using synthetic and real-world data, our developed hardware prototype and its simulator exhibit a high prediction performance and sufficient memory capacity for practical applications, showing promise for future applications in highly integrated AI accelerators.

We developed a three-terminal spintronic memristor named spin-memristor for the artificial synapse of neuromorphic devices. Current-induced domain wall (DW) motion type magnetoresistance was used to generate variable analog conductance changes. Magnetic tunnel junction with a perpendicularly magnetized DW layer was fabricated on a silicon substrate. Due to the forward and backward DW motions, a linear and symmetric conductance response was achieved. The impact of memristive behavior on neural networks was evaluated using numerical simulation of hand-written digit recognition. This spin-memristor is identified as one of the promising candidates for artificial synapses in analog neuromorphic devices. [ABSTRACT FROM AUTHOR]