Topologically Defined Flash Memory

Date of Presentation: 
Wednesday, April 5, 2017
2017 Spring
Research Focus: 

Abstract - Present flash memory uses an analogue quantity, the threshold voltage of a transistor, to store digital data by dividing the available threshold voltage range into two (for SLC), four (MLC), or eight (TLC) levels. Hardwired thresholds are usually assigned to define the boundaries between levels. This hardware defined approach has served the non-volatile memory industry well but faces problems such as reduced program/erase endurance and complexity in dealing with the increasing error correction requirement.
We review alternate methods of storing and retrieving information from flash memory cells to achieve topologically defined flash memory, i.e. without hardwired or fixed threshold between levels. Implementations of these approaches are discussed, including sensing methods where the threshold voltage, as represented by a sensing time, is extracted in one sensing cycle has been proposed. The digital state can then be extract in software. The software defined nature of these memories allows greater flexibility for allocating cells to almost arbitrary digital states and are expected to be more reliable and flexible for future cloud based storage systems.
Bio - Yanjun Ma obtained his Ph.D. degree from the University of Washington in Seattle and is currently a VP of Innovations at the Invention Development Fund (recently renamed Xinova) of Intellectual Ventures. Prior to his current position, Yanjun was a principal engineer and a director of Technology Development and Production Engineering at Impinj, Inc. in Seattle. He also had held senior research and engineering positions at Lattice Semiconductors, Sharp Labs of America as well as research positions at Bell Labs, Brookhaven National Labs, and the University of Washington.
He started his research career in solid state physics. Later, moving to semiconductor industry, he did research on IC technology development, non-volatile memory, low power circuits and RFID systems. Yanjun has 36 issued US patents, including some very highly cited and earliest patents on high k gate dielectrics, and over 80 publications. He is the co-author of “Non-logic Device Design in Logic Processes”.

Yanjun Ma, PhD
Speaker affiliation: 
Intellectual Ventures