Intellectual Property Law

US Patent 7459933 - Pattern recognition using memristor crossbar array

Earlier this year researchers at Hewlett-Packard announced a physical realization of a nanoscale memristor as a fundamentally new circuit element originally theorized by UC Berkeley professor Leon Chua in 1971. One of the benefits of memristors is that they can bridge the gap between software and hardware in the areas of image analysis and pattern recognition. This is one of my basic patent to using hysteretic resistance materials similar to that found for memristors in pattern recognition. I was graciously invited to speak at a symposium on memristors on some of my proposed applications along with Leon Chua and some of the senior scientists at HP. Video of the symposium is available on Youtube here (my portion starts at 1:25:25).

I am currently offering a fairly low licensing deal for my combined patent portfolio related to memristors (currently 6 issued and 9 pending U.S. patents) for early licensees. More details are available here.

Claims 1 and 18 read:

1. A method comprising:

storing two-dimensional binary data in the form of high or low resistance states into a crossbar array with a programmable material layer; and

transforming the two-dimensional binary data into one-dimensional analog data via the crossbar array wherein the stored two-dimensional binary data is image data received from a sensing array.

18. A method comprising:

storing two-dimensional binary data in the form of high or low resistance states into a crossbar array with a programmable material layer;

transforming the two-dimensional binary data into one-dimensional analog data via the crossbar array; and

using the one-dimensional analog data for pattern comparison, pattern location, or pattern tracking.

Full post as published by TinytechIP on December 06, 2008 (boomark / email).