Tag Archives: Felix Messerschmitt

A perovskite memristor with three stable resistive states

Thanks to Dexter Johnson’s Oct. 22, 2015 posting on his Nanoclast blog (on the IEEE [Institute of Electrical and Electronics Engineers]) website, I’ve found information about a second memristor with three terminals, aka, three stable resistive states,  (the first is mentioned in my April 10, 2015 posting). From Dexter’s posting (Note: Links have been removed),

Now researchers at ETH Zurich have designed a memristor device out of perovskite just 5 nanometres thick that has three stable resistive states, which means it can encode data as 0,1 and 2, or a “trit” as opposed to a “bit.”

The research, which was published in the journal ACS Nano, developed model devices that have two competing nonvolatile resistive switching processes. These switching processes can be alternatively triggered by the effective switching voltage and time applied to the device.

“Our component could therefore also be useful for a new type of IT (Information Technology) that is not based on binary logic, but on a logic that provides for information located ‘between’ the 0 and 1,” said Jennifer Rupp, professor in the Department of Materials at ETH Zurich, in a press release. “This has interesting implications for what is referred to as fuzzy logic, which seeks to incorporate a form of uncertainty into the processing of digital information. You could describe it as less rigid computing.”

An Oct. 19, 2015 Swiss National Science Foundation press release provides context for the research,

Two IT giants, Intel and HP, have entered a race to produce a commercial version of memristors, a new electronics component that could one day replace flash memory (DRAM) used in USB memory sticks, SD cards and SSD hard drives. “Basically, memristors require less energy since they work at lower voltages,” explains Jennifer Rupp, professor in the Department of Materials at ETH Zurich and holder of a SNSF professorship grant. “They can be made much smaller than today’s memory modules, and therefore offer much greater density. This means they can store more megabytes of information per square millimetre.” But currently memristors are only at the prototype stage. [emphasis mine]

There is a memristor-based product on the market as I noted in a Sept. 10, 2015 posting, although that may not be the type of memristive device that Rupp seems to be discussing. (Should you have problems accessing the Swiss National Science Foundation press release, you can find a lightly edited version (a brief [two sentences] history of the memristor has been left out) here on Azonano.

Jacopo Prisco wrote for CNN online in a March 2, 2015 article about memristors and Rupp’s work (Note: A link has been removed),

Simply put, the memristor could mean the end of electronics as we know it and the beginning of a new era called “ionics”.

The transistor, developed in 1947, is the main component of computer chips. It functions using a flow of electrons, whereas the memristor couples the electrons with ions, or electrically charged atoms.

In a transistor, once the flow of electrons is interrupted by, say, cutting the power, all information is lost. But a memristor can remember the amount of charge that was flowing through it, and much like a memory stick it will retain the data even when the power is turned off.

This can pave the way for computers that will instantly turn on and off like a light bulb and never lose data: the RAM, or memory, will no longer be erased when the machine is turned off, without the need to save anything to hard drives as with current technology.

Jennifer Rupp is a Professor of electrochemical materials at ETH Zurich, and she’s working with IBM to build a memristor-based machine.

Memristors, she points out, function in a way that is similar to a human brain: “Unlike a transistor, which is based on binary codes, a memristor can have multi-levels. You could have several states, let’s say zero, one half, one quarter, one third, and so on, and that gives us a very powerful new perspective on how our computers may develop in the future,” she told CNN’s Nick Glass.

This is the CNN interview with Rupp,

Prisco also provides an update about HP’s memristor-based product,

After manufacturing the first ever memristor, Hewlett Packard has been working for years on a new type of computer based on the technology. According to plans, it will launch by 2020.

Simply called “The Machine”, it uses “electrons for processing, photons for communication, and ions for storage.”

I first wrote about HP’s The Machine in a June 25, 2014 posting (scroll down about 40% of the way).

There are many academic teams researching memristors including a team at Northwestern University. I highlighted their announcement of a three-terminal version in an April 10, 2015 posting. While Rupp’s team achieved its effect with a perovskite substrate, the Northwestern team used a molybdenum disulfide (MoS2) substrate.

For anyone wanting to read the latest research from ETH, here’s a link to and a citation for the paper,

Uncovering Two Competing Switching Mechanisms for Epitaxial and Ultrathin Strontium Titanate-Based Resistive Switching Bits by Markus Kubicek, Rafael Schmitt, Felix Messerschmitt, and Jennifer L. M. Rupp. ACS Nano, Article ASAP DOI: 10.1021/acsnano.5b02752 Publication Date (Web): October 8, 2015

Copyright © 2015 American Chemical Society

This paper is behind a paywall.

Finally, should you find the commercialization aspects of the memristor story interesting, there’s a June 6, 2015 posting by Knowm CEO (chief executive officer) Alex Nugent waxes eloquent on HP Labs’ ‘memristor problem’ (Note: A link has been removed),

Today I read something that did not surprise me. HP has said that their memristor technology will be replaced by traditional DRAM memory for use in “The Machine”. This is not surprising for those of us who have been in the field since before HP’s memristor marketing engine first revved up in 2008. While I have to admit the miscommunication between HP’s research and business development departments is starting to get really old, I do understand the problem, or at least part of it.

There are two ways to develop memristors. The first way is to force them to behave as you want them to behave. Most memristors that I have seen do not behave like fast, binary, non-volatile, deterministic switches. This is a problem because this is how HP wants them to behave. Consequently a perception has been created that memristors are for non-volatile fast memory. HP wants a drop-in replacement for standard memory because this is a large and established market. Makes sense of course, but its not the whole story on memristors.

Memristors exhibit a huge range of amazing phenomena. Some are very fast to switch but operate probabilistically. Others can be changed a little bit at a time and are ideal for learning. Still others have capacitance (with memory), or act as batteries. I’ve even seen some devices that can be programmed to be a capacitor or a resistor or a memristor. (Seriously).

Nugent, whether you agree with him or not provides, some fascinating insight. In the excerpt I’ve included here, he seems to provide confirmation that it’s possible to state ‘there are no memristors on the market’ and ‘there are memristors on the market’ because different devices are being called memristors.