MIT.nano Dec. Seminar: Neuromorphic Computing via Integration of Diffusive and Drift Memristors

Friday, December 5, 1:00pm - 2:00pm (EST)

Time shown in-04:00 America, New York

J. Joshua Yang
Arthur B. Freeman Chair Professor of Electrical Engineering and Computer Engineering
University of Southern California

Date: December 5, 2025
Time: 1:00 – 2:00 PM ET
Location: Grier 34-401A
Reception to follow.

Memristors can be broadly categorized into diffusive and drift types, distinguished by their reset switching mechanisms. Diffusive memristors reset through spontaneous ion diffusion under zero bias, exhibiting dynamics that closely resemble biological ionic processes, making them particularly attractive for processing temporal information. Drift memristors, in contrast, rely on electric-field-driven ion migration, offering highly stable and tunable analog resistance states that are well suited for storing long-term information. The synergistic integration of diffusive and drift memristors enables the design of compact, energy-efficient, and biorealistic neuromorphic hardware. This talk will present recent progress in device engineering, array-level integration, and application demonstrations, highlighting their potential to advance next-generation computing hardware for machine learning and artificial intelligence.

See the event webpage: https://mitnano.mit.edu/events/dec-seminar-yang

Add to Calendar 2025/12/05 13:00:00 2025/12/05 14:00:00 America/New_York MIT.nano Dec. Seminar: Neuromorphic Computing via Integration of Diffusive and Drift Memristors J. Joshua Yang
Arthur B. Freeman Chair Professor of Electrical Engineering and Computer Engineering
University of Southern California

Date: December 5, 2025
Time: 1:00 – 2:00 PM ET
Location: Grier 34-401A
Reception to follow.

Memristors can be broadly categorized into diffusive and drift types, distinguished by their reset switching mechanisms. Diffusive memristors reset through spontaneous ion diffusion under zero bias, exhibiting dynamics that closely resemble biological ionic processes, making them particularly attractive for processing temporal information. Drift memristors, in contrast, rely on electric-field-driven ion migration, offering highly stable and tunable analog resistance states that are well suited for storing long-term information. The synergistic integration of diffusive and drift memristors enables the design of compact, energy-efficient, and biorealistic neuromorphic hardware. This talk will present recent progress in device engineering, array-level integration, and application demonstrations, highlighting their potential to advance next-generation computing hardware for machine learning and artificial intelligence.

See the event webpage: https://mitnano.mit.edu/events/dec-seminar-yang 34-401A false MM/DD/YYYY 30 OPAQUE ecc7a70de8bc40a0861fa5ec627d9524

34-401A

MITnano, mitnano@mit.edu

Link: