"Self-Adaptable Memristive Memories for Optimal Detection Performance"

Sangho Shin, Post-doctoral Researcher, University of California, Merced

October 12th (Tuesday), 1:00pm
Harold Frank Hall 4164

There have been world-wide efforts to create new memories based on resistance changes induced by ionic motion in various materials. In this talk, a simple passive memristive memory model that can handle any random distribution of stored data-patterns will be introduced dealing with sneak currents and optimal detection performances. Statistical approximations of a complex set of multi nodal equations that are based on traditional circuit analysis methods are done to build a 2×2 equivalent statistical model for a generalized nxm passive memristive memory array. The statistical model significantly improves the computational efficiency enabling a broad range of analysis, such as the data-pattern dependant optimum detection performance, with no limit on the memory array size, and thus helps develop reliable memory architectures and circuits for timely productions. In particular, considering that the circuit parameters are desirably to be adapted to maximize the readout performances for all possible stored data-patterns, viable memory architectures and circuit configurations with self-adaptability to its statistically desired optimal value will be presented. The self-adaptable memories are expected to significantly enhance the detection performance along with reduced power consumption, compared to the conventional non-adaptable memories.

This talk will also include brief introductions to our recent research activities related to the memristive devices and systems, such as programmable analog ICs and field programmable stateful logic arrays.

About Sangho Shin, Post-doctoral Researcher:

Sangho Shin received his M.S. and Ph. D. degrees from the Dept. of Electrical Engineering, Korea Advanced Institute of Science and Technology (KAIST), Daejeon, Korea, respectively in 2002 and 2007. He is currently working as a post-doctoral researcher with the School of Engineering, University of California, Merced, where he is in charge of running a research group with broad activities on the elusive memristive devices and systems exploiting their models and useful applications. His research interests include low-power analog/RF integrated circuits; low-power very large scale integration design; mixed-signal mixed-technology integrated systems; modeling and simulation of semiconductor devices and circuits; high-speed input/output schemes; analysis and estimation of 3-D thermal distribution.

Hosted by: Professor Dmitri Strukov