Knol on Memristor Electronics

Recently I had another article rejected from IEEE Transactions on Nanotechnology. Some of the peer reviewers provided feedback noting that I did not cite enough prior references and suggested that I should use a circuit simulation tool to verify the operability of the design.

However, Google's recent release of their "Knol" project gives me an alternative way to publish and maybe attract some attention to my circuit design ideas. I'm working on a new article on applications of memristors here which is related to some of my patents. While of course not as prestigious as an IEEE article at least it gives me an opportunity to present my ideas in a forum for others to easily access.

Fourth Patent Issued - US 7,391,235

My fourth patent issued this week directed to a programmable resistance switch crossbar circuit combined with operational amplifier circuitry useful to applications such as programmable waveforms, reconfigurable filters, and communication switching.

Some of my other patent applications have had final rejections issued by the PTO. US2007/0257704 is directed to a method of reconfiguring a programmable crossbar signal processor in which the claims were amended during prosecution to distinguish the claimed method from reconfiguration of programmble logic arrays as performed in the prior art. However, the Examiner made the rejection final with prior art directed to programmble logic (in contrast to the claim language) so I may have to appeal this one. US2007/0233761, on the other hand, is directed to a programmable crossbar array used in an arithmetic processor. The Examiner made a rejection based on an optical crossbar switching array used for arithmetic processing that is alleged to be obviously convertible to a resistance switching crossbar array (despite that the reference appears to teach away from electronic based configurations). This one is under appeal.

While I'm still trying to get some papers published discussing my technology most of the journals I've submitted to have either rejected the submission for being outside the scope of the journal or for not having sufficient scientific merit since no experimental or computational result is included (the papers were directed to a new technology not a new measurement). One exception was the 2008 NSTI conference which published a paper I wrote on logicless computation with crossbar arrays.

Recently I was contacted by a representative of Ocean Tomo about auctioning one of my patents (US 7,342,413) related to scanning probe interconnects for nanowire crossbar arrays which may have some benefits to ultrahigh density non-volatile memory. Although I have no interest in this approach it is good to know that some of my patents are attracting attention. Hopefully, companies interested in licensing my patents will find similar value in my technology as I spread the word.

Patent Licensing Information

In recent years there has been an acceleration in the development of new materials used in electronics design. Much of the focus has been on nanomaterials due to the trend of scale reduction predicted by Moore's Law. However, another area of interest is in reconfigurable materials which are capable of changing state based on an applied threshold voltage or current. So far the two main proposals for applications attempting to exploit such materials are non-volatile memory (RRAM) and some programmable logic designs but there are a variety of other application areas such as signal processing, pattern recognition, control systems, and programmable waveform circuitry which may also benefit by more adaptive designs using these materials. The development of "memristor" material by HP has drawn increased attention to the potential of such applications. In addition, as described in my paper "Logicless Computational Architectures with Nanoscale Crossbar Arrays" the use of resistance switching materials in crossbar arrays connected to analog processing circuitry offers an alternative computational paradigm with advantages to large number computation and parallel processing capabilities.

Over the past few years I have developed a variety of circuit designs based on the exploitation of resistance switching materials for signal processing, pattern recognition, control systems, arithmetic processing, and programmable waveforms. For those interested in my technology I am currently offering a licensing deal covering all of my patents based on resistance switching materials used in the noted applications. As an incentive to attract companies willing to invest at this early stage the license deal includes a 4% running royalty that can optionally be converted to a one-time paid-up license. Thus if a product is developed based on my patents which is seen to be particularly valuable the license may be bought out without further royalty payments. However, this option will only be available to the first few companies interested in development and commercialization of the technology.

For information on particular ones of my issued or pending patents see the following links:

Issued Patents

Pending Patent Applications

For further information I can be contacted at:

Blaise Mouttet
4201 Wilson Blvd. 110-364
Arlington, Va 22203
(571) 345-8890
blaise_mouttet@yahoo.com

Non-final rejection (11/418,057) and allowance (11/790,713)

I received a non-final rejection this week from my patent application 11/418,057 directed to an electron beam nanotip array configured for lithography. Conventionally while electron beam lithography has advantages over optical lithography in terms of resolution it is a slow process making it inconvenient for large scale applications. My patent application teaches forming a high resolution vertical array of nanotips (such as carbon nanotubes) operated in parallel that can be digitally controlled to pattern a target more quickly. Similar electron emitting CNT arrays have been applied in the prior art to field emission displays and the Examner made a rejection based on some of these references. However, slight modification of the claims should be able to overcome the rejections.

Another of my patent application 11/790,713 directed to programmable impedance material used in a signal processor was also allowed this week. This patent may have some relevance to future applications based on memristor and similar materials.

Memristor - the key to strong A.I. ?

Hewlett Packard's Information and Quantum Systems Lab recently reported on the physical realization of a new circuit element that combines the aspects of a resistor and a memory device called a memristor (see EETimes article). The "memristor" developed by HP is noted as a dual layer of titanium dioxide thin films that possesses hysteretic behavior so that the resistance can by switched by an order of magnitude of 1000 and which possesses qualities making it analogous to a neuron. Although thin film oxides used in variable resistance materials have been previously reported, they have primarily been suggested for non-volatile memory applications rather than for signal processing or pattern recognition applications. However, my various patents and patent applications teach a variety of electronic circuit configurations that implement resistance variable materials used in crossbars to construct pattern recognition circuitry, programmable control systems, waveform generation devices, arithmetic processing circuitry, and a variety of other applications.

Also of interest is this patent application from Samsung which appears to disclose the dual TiO2 resistance switching material and effect.

Logicless Computational Architectures using Nanoscale Crossbar Arrays

So far my efforts at attracting companies interested in developing products based on my inventions have not gone well. While my professors at George Mason University have offered some positive feedback on some of my ideas the "not invented here" syndrome seems alive and well in many corporations. My current strategy is to publish some technical papers based on my ideas and use these papers to attract interest and corporate sponsorship in further development of the technology I am proposing. My first effort in this direction is a poster that I am presenting at the upcoming 2008 NSTI conference.

The poster describes a new type of computational system based on resistance switching crossbars. Greg Snider of Hewlett-Packard and Nantero have already made some suggestion for using crossbars for computational purposes but these suggestions have so far been limited to attempts at reproducing basic logic gates. The problem with that approach is that to create full-adders and more complex arithmetic systems multiple "tiles" of the crossbars need to be interconnected with each tile only performing the function of a single logic gate. The approach discussed in my paper instead uses a single crossbar to perform 4 bit addition using a hybrid analog/digital approach. In addition, my approach may be more easily integrable with microscale electronic components and provide some advantages in integrating data storage functions with data processing functions. Hopefully this paper and future papers that I am currently attempting to get published will at least give me some more credibility when attempting to attract companies to commercialize my ideas.

Notice of Allowance - U.S. Application 11/790,712

Another of my continuations received a notice of allowance from the PTO on 3/27/2008. This patent focuses on a symmetrical crossbar structure with dual rectification layers to make reprogramming of resistance switching material easier.