Over the past decade, the rapid development of information and communication technologies opened new horizons for artificial intelligence and green technologies. These developments challenge the state-of-the-art of nanoelectronics in demanding new hardware architectures to overcome von Neumann computing by implementing neuromorphic-type of data processing. Memristive devices have recently gained tremendous interest not only in the scientific community but also in the semiconductor industry as building blocks for hardware implementation of in-memory computing and artificial intelligence, due to their ability to emulate neuromorphic type of data processing combined with high scalability down to almost atomic scale, low power consumption (<pJ for operation) and high operational speed (ps).
The revision of the SI in 2019 represents an historic change of paradigm for metrology, defining all the SI units in terms of fundamental constants of nature. The development of new experiments and devices are now needed for correlating physical observables to the fixed defining constants, paving the way for the realization of self-calibrating systems that can independently refer to the fundamental constants of nature with zero-chain traceability. In accordance with the revised SI, memristive devices exhibiting quantized conductance levels represent promising platforms for on-chip integrated and complementary metal-oxide-semiconductor (CMOS) compatible standard of resistance, which work in air and at room temperature.
While novel materials and devices for nanoelectronics indeed offer many potential benefits, they also bring challenges for testing and characterization. As an emerging technology, memristive devices lack standardization and insights in the fundamental physics underpinning its technology, hindering their further development. Understanding and controlling resistive switching behavior at the nanoscale is therefore highly challenging and high throughput metrology is urgently required. The development of new technologies for nanoelectronics, including quantum technologies, neuromorphic computing and related metrology, is in line with the European R&D programmes “Quantum Technologies Flagship” and the “Human Brain Flagship”. For this purpose, memristive model systems need to be developed, to establish a relationship in between material properties and device functionalities.
The understanding of nanoionic processes involved in memristive devices requires i) advancements in nanoelectrical characterization techniques, ii) the development of a traceable quantification of chemical, structural and ionic/electronic properties of memristive devices and iii) the development of metrological cross-platforms measurement techniques. These are the key requirements for understanding and controlling quantized conductance effects in such devices for the realization of the standard of resistance.