Electrical Characterization Equipment for Memristive Devices

Radio Frequency Measurement Setup

The radio frequency setup allows the electrical characterization of electronic devices at frequencies up to 40 GHz. Using coplanar waveguide structures (CPW), voltage pulses on a sub-nanosecond time scale can be delivered to the device under test. It is used to characterize the switching kinetics of memristive devices and to determine their switching speed limits. With this setup, the fastest measured writing time of a memristive device was measured which amounts to only 50 ps [1]. 

AI-Matrix-Tester

The custom test system provided by aixACCT Systems combines a semi-automatic 200 mm prober with electrical characterization equipment for two operation modes: (i) Single device characterization via (2 or 4 probe) wedge contacting, 4 DAC and 4 ADC (both with 1 GS/s, 16 bit). (ii) Matrix characterization with 64 probes (up to 32 x 32 device arrays), 64 DAC (200 MS/s, 16 bit) and 32 ADC (100 MS/s, 16 bit). Both modes are equipped with a fast current compliance implemented close to the probes. 

The setup is used for the automated electrical characterization of memristive cells and also allows the biasing of 1-transistor-1-memristor (1T1R) devices. Using the matrix contacting mode (ii), computation in memory (CIM) and neuromorphic computing approaches can be tested on arrays of memristive elements. It is anticipated that neuromorphic computing based on memristive elements establishes the next generation of energy efficient computing for artificial intelligence applications. Research and development in this field is conducted in the NEUROTEC project, in which NEUROTEC stands for neuroinspired technologies for the artificial intelligence (AI) of future electronics in the Rhineland, as well as in the Cluster for future NeuroSys, where NeuroSys stands for Neuromorphe Hardware for autonomous Systems.

[1] von Witzleben, M., et al. "Study of the SET switching event of VCM-based memories on a picosecond timescale." Journal of applied physics 127.20 (2020): 204501

Detailed information on Memristive Devices can be found on the Projectwebsite