Category Archives: neuromorphic engineering

Less is more—a superconducting synapse

It seems the US National Institute of Standards and Technology (NIST) is more deeply invested into developing artificial brains than I had realized (See: April 17, 2018 posting). A January 26, 2018 NIST news release on EurekAlert describes the organization’s latest foray into the field,

Researchers at the National Institute of Standards and Technology (NIST) have built a superconducting switch that “learns” like a biological system and could connect processors and store memories in future computers operating like the human brain.

The NIST switch, described in Science Advances, is called a synapse, like its biological counterpart, and it supplies a missing piece for so-called neuromorphic computers. Envisioned as a new type of artificial intelligence, such computers could boost perception and decision-making for applications such as self-driving cars and cancer diagnosis.

A synapse is a connection or switch between two brain cells. NIST’s artificial synapse–a squat metallic cylinder 10 micrometers in diameter–is like the real thing because it can process incoming electrical spikes to customize spiking output signals. This processing is based on a flexible internal design that can be tuned by experience or its environment. The more firing between cells or processors, the stronger the connection. Both the real and artificial synapses can thus maintain old circuits and create new ones. Even better than the real thing, the NIST synapse can fire much faster than the human brain–1 billion times per second, compared to a brain cell’s 50 times per second–using just a whiff of energy, about one ten-thousandth as much as a human synapse. In technical terms, the spiking energy is less than 1 attojoule, lower than the background energy at room temperature and on a par with the chemical energy bonding two atoms in a molecule.

“The NIST synapse has lower energy needs than the human synapse, and we don’t know of any other artificial synapse that uses less energy,” NIST physicist Mike Schneider said.

The new synapse would be used in neuromorphic computers made of superconducting components, which can transmit electricity without resistance, and therefore, would be more efficient than other designs based on semiconductors or software. Data would be transmitted, processed and stored in units of magnetic flux. Superconducting devices mimicking brain cells and transmission lines have been developed, but until now, efficient synapses–a crucial piece–have been missing.

The brain is especially powerful for tasks like context recognition because it processes data both in sequence and simultaneously and stores memories in synapses all over the system. A conventional computer processes data only in sequence and stores memory in a separate unit.

The NIST synapse is a Josephson junction, long used in NIST voltage standards. These junctions are a sandwich of superconducting materials with an insulator as a filling. When an electrical current through the junction exceeds a level called the critical current, voltage spikes are produced. The synapse uses standard niobium electrodes but has a unique filling made of nanoscale clusters of manganese in a silicon matrix.

The nanoclusters–about 20,000 per square micrometer–act like tiny bar magnets with “spins” that can be oriented either randomly or in a coordinated manner.

“These are customized Josephson junctions,” Schneider said. “We can control the number of nanoclusters pointing in the same direction, which affects the superconducting properties of the junction.”

The synapse rests in a superconducting state, except when it’s activated by incoming current and starts producing voltage spikes. Researchers apply current pulses in a magnetic field to boost the magnetic ordering, that is, the number of nanoclusters pointing in the same direction. This magnetic effect progressively reduces the critical current level, making it easier to create a normal conductor and produce voltage spikes.

The critical current is the lowest when all the nanoclusters are aligned. The process is also reversible: Pulses are applied without a magnetic field to reduce the magnetic ordering and raise the critical current. This design, in which different inputs alter the spin alignment and resulting output signals, is similar to how the brain operates.

Synapse behavior can also be tuned by changing how the device is made and its operating temperature. By making the nanoclusters smaller, researchers can reduce the pulse energy needed to raise or lower the magnetic order of the device. Raising the operating temperature slightly from minus 271.15 degrees C (minus 456.07 degrees F) to minus 269.15 degrees C (minus 452.47 degrees F), for example, results in more and higher voltage spikes.

Crucially, the synapses can be stacked in three dimensions (3-D) to make large systems that could be used for computing. NIST researchers created a circuit model to simulate how such a system would operate.

The NIST synapse’s combination of small size, superfast spiking signals, low energy needs and 3-D stacking capability could provide the means for a far more complex neuromorphic system than has been demonstrated with other technologies, according to the paper.

NIST has prepared an animation illustrating the research,

Caption: This is an animation of how NIST’s artificial synapse works. Credit: Sean Kelley/NIST

Here’s a link to and a citation for the paper,

Ultralow power artificial synapses using nanotextured magnetic Josephson junctions by Michael L. Schneider, Christine A. Donnelly, Stephen E. Russek, Burm Baek, Matthew R. Pufall, Peter F. Hopkins, Paul D. Dresselhaus, Samuel P. Benz, and William H. Rippard. Science Advances 26 Jan 2018: Vol. 4, no. 1, e1701329 DOI: 10.1126/sciadv.1701329

This paper is open access.

Samuel K. Moore in a January 26, 2018 posting on the Nanoclast blog (on the IEEE [Institute for Electrical and Electronics Engineers] website) describes the research and adds a few technical explanations such as this about the Josephson junction,

In a magnetic Josephson junction, that “weak link” is magnetic. The higher the magnetic field, the lower the critical current needed to produce voltage spikes. In the device Schneider and his colleagues designed, the magnetic field is caused by 20,000 or so nanometer-scale clusters of manganese embedded in silicon. …

Moore also provides some additional links including this one to his November 29, 2017 posting where he describes four new approaches to computing including quantum computing and neuromorphic (brain-like) computing.

New path to viable memristor/neuristor?

I first stumbled onto memristors and the possibility of brain-like computing sometime in 2008 (around the time that R. Stanley Williams and his team at HP Labs first published the results of their research linking Dr. Leon Chua’s memristor theory to their attempts to shrink computer chips). In the almost 10 years since, scientists have worked hard to utilize memristors in the field of neuromorphic (brain-like) engineering/computing.

A January 22, 2018 news item on phys.org describes the latest work,

When it comes to processing power, the human brain just can’t be beat.

Packed within the squishy, football-sized organ are somewhere around 100 billion neurons. At any given moment, a single neuron can relay instructions to thousands of other neurons via synapses—the spaces between neurons, across which neurotransmitters are exchanged. There are more than 100 trillion synapses that mediate neuron signaling in the brain, strengthening some connections while pruning others, in a process that enables the brain to recognize patterns, remember facts, and carry out other learning tasks, at lightning speeds.

Researchers in the emerging field of “neuromorphic computing” have attempted to design computer chips that work like the human brain. Instead of carrying out computations based on binary, on/off signaling, like digital chips do today, the elements of a “brain on a chip” would work in an analog fashion, exchanging a gradient of signals, or “weights,” much like neurons that activate in various ways depending on the type and number of ions that flow across a synapse.

In this way, small neuromorphic chips could, like the brain, efficiently process millions of streams of parallel computations that are currently only possible with large banks of supercomputers. But one significant hangup on the way to such portable artificial intelligence has been the neural synapse, which has been particularly tricky to reproduce in hardware.

Now engineers at MIT [Massachusetts Institute of Technology] have designed an artificial synapse in such a way that they can precisely control the strength of an electric current flowing across it, similar to the way ions flow between neurons. The team has built a small chip with artificial synapses, made from silicon germanium. In simulations, the researchers found that the chip and its synapses could be used to recognize samples of handwriting, with 95 percent accuracy.

A January 22, 2018 MIT news release by Jennifer Chua (also on EurekAlert), which originated the news item, provides more detail about the research,

The design, published today [January 22, 2018] in the journal Nature Materials, is a major step toward building portable, low-power neuromorphic chips for use in pattern recognition and other learning tasks.

The research was led by Jeehwan Kim, the Class of 1947 Career Development Assistant Professor in the departments of Mechanical Engineering and Materials Science and Engineering, and a principal investigator in MIT’s Research Laboratory of Electronics and Microsystems Technology Laboratories. His co-authors are Shinhyun Choi (first author), Scott Tan (co-first author), Zefan Li, Yunjo Kim, Chanyeol Choi, and Hanwool Yeon of MIT, along with Pai-Yu Chen and Shimeng Yu of Arizona State University.

Too many paths

Most neuromorphic chip designs attempt to emulate the synaptic connection between neurons using two conductive layers separated by a “switching medium,” or synapse-like space. When a voltage is applied, ions should move in the switching medium to create conductive filaments, similarly to how the “weight” of a synapse changes.

But it’s been difficult to control the flow of ions in existing designs. Kim says that’s because most switching mediums, made of amorphous materials, have unlimited possible paths through which ions can travel — a bit like Pachinko, a mechanical arcade game that funnels small steel balls down through a series of pins and levers, which act to either divert or direct the balls out of the machine.

Like Pachinko, existing switching mediums contain multiple paths that make it difficult to predict where ions will make it through. Kim says that can create unwanted nonuniformity in a synapse’s performance.

“Once you apply some voltage to represent some data with your artificial neuron, you have to erase and be able to write it again in the exact same way,” Kim says. “But in an amorphous solid, when you write again, the ions go in different directions because there are lots of defects. This stream is changing, and it’s hard to control. That’s the biggest problem — nonuniformity of the artificial synapse.”

A perfect mismatch

Instead of using amorphous materials as an artificial synapse, Kim and his colleagues looked to single-crystalline silicon, a defect-free conducting material made from atoms arranged in a continuously ordered alignment. The team sought to create a precise, one-dimensional line defect, or dislocation, through the silicon, through which ions could predictably flow.

To do so, the researchers started with a wafer of silicon, resembling, at microscopic resolution, a chicken-wire pattern. They then grew a similar pattern of silicon germanium — a material also used commonly in transistors — on top of the silicon wafer. Silicon germanium’s lattice is slightly larger than that of silicon, and Kim found that together, the two perfectly mismatched materials can form a funnel-like dislocation, creating a single path through which ions can flow.

The researchers fabricated a neuromorphic chip consisting of artificial synapses made from silicon germanium, each synapse measuring about 25 nanometers across. They applied voltage to each synapse and found that all synapses exhibited more or less the same current, or flow of ions, with about a 4 percent variation between synapses — a much more uniform performance compared with synapses made from amorphous material.

They also tested a single synapse over multiple trials, applying the same voltage over 700 cycles, and found the synapse exhibited the same current, with just 1 percent variation from cycle to cycle.

“This is the most uniform device we could achieve, which is the key to demonstrating artificial neural networks,” Kim says.

Writing, recognized

As a final test, Kim’s team explored how its device would perform if it were to carry out actual learning tasks — specifically, recognizing samples of handwriting, which researchers consider to be a first practical test for neuromorphic chips. Such chips would consist of “input/hidden/output neurons,” each connected to other “neurons” via filament-based artificial synapses.

Scientists believe such stacks of neural nets can be made to “learn.” For instance, when fed an input that is a handwritten ‘1,’ with an output that labels it as ‘1,’ certain output neurons will be activated by input neurons and weights from an artificial synapse. When more examples of handwritten ‘1s’ are fed into the same chip, the same output neurons may be activated when they sense similar features between different samples of the same letter, thus “learning” in a fashion similar to what the brain does.

Kim and his colleagues ran a computer simulation of an artificial neural network consisting of three sheets of neural layers connected via two layers of artificial synapses, the properties of which they based on measurements from their actual neuromorphic chip. They fed into their simulation tens of thousands of samples from a handwritten recognition dataset commonly used by neuromorphic designers, and found that their neural network hardware recognized handwritten samples 95 percent of the time, compared to the 97 percent accuracy of existing software algorithms.

The team is in the process of fabricating a working neuromorphic chip that can carry out handwriting-recognition tasks, not in simulation but in reality. Looking beyond handwriting, Kim says the team’s artificial synapse design will enable much smaller, portable neural network devices that can perform complex computations that currently are only possible with large supercomputers.

“Ultimately we want a chip as big as a fingernail to replace one big supercomputer,” Kim says. “This opens a stepping stone to produce real artificial hardware.”

This research was supported in part by the National Science Foundation.

Here’s a link to and a citation for the paper,

SiGe epitaxial memory for neuromorphic computing with reproducible high performance based on engineered dislocations by Shinhyun Choi, Scott H. Tan, Zefan Li, Yunjo Kim, Chanyeol Choi, Pai-Yu Chen, Hanwool Yeon, Shimeng Yu, & Jeehwan Kim. Nature Materials (2018) doi:10.1038/s41563-017-0001-5 Published online: 22 January 2018

This paper is behind a paywall.

For the curious I have included a number of links to recent ‘memristor’ postings here,

January 22, 2018: Memristors at Masdar

January 3, 2018: Mott memristor

August 24, 2017: Neuristors and brainlike computing

June 28, 2017: Dr. Wei Lu and bio-inspired ‘memristor’ chips

May 2, 2017: Predicting how a memristor functions

December 30, 2016: Changing synaptic connectivity with a memristor

December 5, 2016: The memristor as computing device

November 1, 2016: The memristor as the ‘missing link’ in bioelectronic medicine?

You can find more by using ‘memristor’ as the search term in the blog search function or on the search engine of your choice.

Thanks for the memory: the US National Institute of Standards and Technology (NIST) and memristors

In January 2018 it seemed like I was tripping across a lot of memristor stories . This came from a January 19, 2018 news item on Nanowerk,

In the race to build a computer that mimics the massive computational power of the human brain, researchers are increasingly turning to memristors, which can vary their electrical resistance based on the memory of past activity. Scientists at the National Institute of Standards and Technology (NIST) have now unveiled the long-mysterious inner workings of these semiconductor elements, which can act like the short-term memory of nerve cells.

A January 18, 2018 NIST news release (also on EurekAlert), which originated the news item, fills in the details,

Just as the ability of one nerve cell to signal another depends on how often the cells have communicated in the recent past, the resistance of a memristor depends on the amount of current that recently flowed through it. Moreover, a memristor retains that memory even when electrical power is switched off.

But despite the keen interest in memristors, scientists have lacked a detailed understanding of how these devices work and have yet to develop a standard toolset to study them.

Now, NIST scientists have identified such a toolset and used it to more deeply probe how memristors operate. Their findings could lead to more efficient operation of the devices and suggest ways to minimize the leakage of current.

Brian Hoskins of NIST and the University of California, Santa Barbara, along with NIST scientists Nikolai Zhitenev, Andrei Kolmakov, Jabez McClelland and their colleagues from the University of Maryland’s NanoCenter (link is external) in College Park and the Institute for Research and Development in Microtechnologies in Bucharest, reported the findings (link is external) in a recent Nature Communications.

To explore the electrical function of memristors, the team aimed a tightly focused beam of electrons at different locations on a titanium dioxide memristor. The beam knocked free some of the device’s electrons, which formed ultrasharp images of those locations. The beam also induced four distinct currents to flow within the device. The team determined that the currents are associated with the multiple interfaces between materials in the memristor, which consists of two metal (conducting) layers separated by an insulator.

“We know exactly where each of the currents are coming from because we are controlling the location of the beam that is inducing those currents,” said Hoskins.

In imaging the device, the team found several dark spots—regions of enhanced conductivity—which indicated places where current might leak out of the memristor during its normal operation. These leakage pathways resided outside the memristor’s core—where it switches between the low and high resistance levels that are useful in an electronic device. The finding suggests that reducing the size of a memristor could minimize or even eliminate some of the unwanted current pathways. Although researchers had suspected that might be the case, they had lacked experimental guidance about just how much to reduce the size of the device.

Because the leakage pathways are tiny, involving distances of only 100 to 300 nanometers, “you’re probably not going to start seeing some really big improvements until you reduce dimensions of the memristor on that scale,” Hoskins said.

To their surprise, the team also found that the current that correlated with the memristor’s switch in resistance didn’t come from the active switching material at all, but the metal layer above it. The most important lesson of the memristor study, Hoskins noted, “is that you can’t just worry about the resistive switch, the switching spot itself, you have to worry about everything around it.” The team’s study, he added, “is a way of generating much stronger intuition about what might be a good way to engineer memristors.”

Here’s a link to and a citation for the paper,

Stateful characterization of resistive switching TiO2 with electron beam induced currents by Brian D. Hoskins, Gina C. Adam, Evgheni Strelcov, Nikolai Zhitenev, Andrei Kolmakov, Dmitri B. Strukov, & Jabez J. McClelland. Nature Communications 8, Article number: 1972 (2017) doi:10.1038/s41467-017-02116-9 Published online: 07 December 2017

This is an open access paper.

It might be my imagination but it seemed like a lot of papers from 2017 were being publicized in early 2018.

Finally, I borrowed much of my headline from the NIST’s headline for its news release, specifically, “Thanks for the memory,” which is a rather old song,

Bob Hope and Shirley Ross in “The Big Broadcast of 1938.”

New breed of memristors?

This new ‘breed’ of memristor (a component in brain-like/neuromorphic computing) is a kind of thin film. First, here’s an explanation of neuromorphic computing from the Finnish researchers looking into a new kind of memristor, from a January 10, 2018 news item on Nanowerk,

The internet of things [IOT] is coming, that much we know. But still it won’t; not until we have components and chips that can handle the explosion of data that comes with IoT. In 2020, there will already be 50 billion industrial internet sensors in place all around us. A single autonomous device – a smart watch, a cleaning robot, or a driverless car – can produce gigabytes of data each day, whereas an airbus may have over 10 000 sensors in one wing alone.

Two hurdles need to be overcome. First, current transistors in computer chips must be miniaturized to the size of only few nanometres; the problem is they won’t work anymore then. Second, analysing and storing unprecedented amounts of data will require equally huge amounts of energy. Sayani Majumdar, Academy Fellow at Aalto University, along with her colleagues, is designing technology to tackle both issues.

Majumdar has with her colleagues designed and fabricated the basic building blocks of future components in what are called “neuromorphic” computers inspired by the human brain. It’s a field of research on which the largest ICT companies in the world and also the EU are investing heavily. Still, no one has yet come up with a nano-scale hardware architecture that could be scaled to industrial manufacture and use.

An Aalto University January 10, 2018 press release, which originated the news item, provides more detail about the work,

“The technology and design of neuromorphic computing is advancing more rapidly than its rival revolution, quantum computing. There is already wide speculation both in academia and company R&D about ways to inscribe heavy computing capabilities in the hardware of smart phones, tablets and laptops. The key is to achieve the extreme energy-efficiency of a biological brain and mimic the way neural networks process information through electric impulses,” explains Majumdar.

Basic components for computers that work like the brain

In their recent article in Advanced Functional Materials, Majumdar and her team show how they have fabricated a new breed of “ferroelectric tunnel junctions”, that is, few-nanometre-thick ferroelectric thin films sandwiched between two electrodes. They have abilities beyond existing technologies and bode well for energy-efficient and stable neuromorphic computing.

The junctions work in low voltages of less than five volts and with a variety of electrode materials – including silicon used in chips in most of our electronics. They also can retain data for more than 10 years without power and be manufactured in normal conditions.

Tunnel junctions have up to this point mostly been made of metal oxides and require 700 degree Celsius temperatures and high vacuums to manufacture. Ferroelectric materials also contain lead which makes them – and all our computers – a serious environmental hazard.

“Our junctions are made out of organic hydro-carbon materials and they would reduce the amount of toxic heavy metal waste in electronics. We can also make thousands of junctions a day in room temperature without them suffering from the water or oxygen in the air”, explains Majumdar.

What makes ferroelectric thin film components great for neuromorphic computers is their ability to switch between not only binary states – 0 and 1 – but a large number of intermediate states as well. This allows them to ‘memorise’ information not unlike the brain: to store it for a long time with minute amounts of energy and to retain the information they have once received – even after being switched off and on again.

We are no longer talking of transistors, but ‘memristors’. They are ideal for computation similar to that in biological brains.  Take for example the Mars 2020 Rover about to go chart the composition of another planet. For the Rover to work and process data on its own using only a single solar panel as an energy source, the unsupervised algorithms in it will need to use an artificial brain in the hardware.

“What we are striving for now, is to integrate millions of our tunnel junction memristors into a network on a one square centimetre area. We can expect to pack so many in such a small space because we have now achieved a record-high difference in the current between on and off-states in the junctions and that provides functional stability. The memristors could then perform complex tasks like image and pattern recognition and make decisions autonomously,” says Majumdar.

The probe-station device (the full instrument, left, and a closer view of the device connection, right) which measures the electrical responses of the basic components for computers mimicking the human brain. The tunnel junctions are on a thin film on the substrate plate. Photo: Tapio Reinekoski

Here’s a link to and a citation for the paper,

Electrode Dependence of Tunneling Electroresistance and Switching Stability in Organic Ferroelectric P(VDF-TrFE)-Based Tunnel Junctions by Sayani Majumdar, Binbin Chen, Qi Hang Qin, Himadri S. Majumdar, and Sebastiaan van Dijken. Advanced Functional Materials Vol. 28 Issue 2 DOI: 10.1002/adfm.201703273 Version of Record online: 27 NOV 2017

© 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

This paper is behind a paywall.

Brain-like computing and memory with magnetoresistance

This is an approach to brain-like computing that’s new (to me, anyway). From a January 9, 2018 news item on Nanowerk (Note: A link has been removed),

From various magnetic tapes, floppy disks and computer hard disk drives, magnetic materials have been storing our electronic information along with our valuable knowledge and memories for well over half of a century.

In more recent years, the new types [sic] phenomena known as magnetoresistance, which is the tendency of a material to change its electrical resistance when an externally-applied magnetic field or its own magnetization is changed, has found its success in hard disk drive read heads, magnetic field sensors and the rising star in the memory technologies, the magnetoresistive random access memory.

A new discovery, led by researchers at the University of Minnesota, demonstrates the existence of a new kind of magnetoresistance involving topological insulators that could result in improvements in future computing and computer storage. The details of their research are published in the most recent issue of the scientific journal Nature Communications (“Unidirectional spin-Hall and Rashba-Edelstein magnetoresistance in topological insulator-ferromagnet layer heterostructures”).

This image illustrates the work,

The schematic figure illustrates the concept and behavior of magnetoresistance. The spins are generated in topological insulators. Those at the interface between ferromagnet and topological insulators interact with the ferromagnet and result in either high or low resistance of the device, depending on the relative directions of magnetization and spins. Credit: University of Minnesota

A January 9, 2018 University of Minnesota College of Science and Engineering news release, which originated the news item, expands on the theme,

“Our discovery is one missing piece of the puzzle to improve the future of low-power computing and memory for the semiconductor industry, including brain-like computing and chips for robots and 3D magnetic memory,” said University of Minnesota Robert F. Hartmann Professor of Electrical and Computer Engineering Jian-Ping Wang, director of the Center for Spintronic Materials, Interfaces, and Novel Structures (C-SPIN) based at the University of Minnesota and co-author of the study.

Emerging technology using topological insulators

While magnetic recording still dominates data storage applications, the magnetoresistive random access memory is gradually finding its place in the field of computing memory. From the outside, they are unlike the hard disk drives which have mechanically spinning disks and swinging heads—they are more like any other type of memory. They are chips (solid state) which you’d find being soldered on circuit boards in a computer or mobile device.

Recently, a group of materials called topological insulators has been found to further improve the writing energy efficiency of magnetoresistive random access memory cells in electronics. However, the new device geometry demands a new magnetoresistance phenomenon to accomplish the read function of the memory cell in 3D system and network.

Following the recent discovery of the unidirectional spin Hall magnetoresistance in a conventional metal bilayer material systems, researchers at the University of Minnesota collaborated with colleagues at Pennsylvania State University and demonstrated for the first time the existence of such magnetoresistance in the topological insulator-ferromagnet bilayers.

The study confirms the existence of such unidirectional magnetoresistance and reveals that the adoption of topological insulators, compared to heavy metals, doubles the magnetoresistance performance at 150 Kelvin (-123.15 Celsius). From an application perspective, this work provides the missing piece of the puzzle to create a proposed 3D and cross-bar type computing and memory device involving topological insulators by adding the previously missing or very inconvenient read functionality.

In addition to Wang, researchers involved in this study include Yang Lv, Delin Zhang and Mahdi Jamali from the University of Minnesota Department of Electrical and Computer Engineering and James Kally, Joon Sue Lee and Nitin Samarth from Pennsylvania State University Department of Physics.

This research was funded by the Center for Spintronic Materials, Interfaces and Novel Architectures (C-SPIN) at the University of Minnesota, a Semiconductor Research Corporation program sponsored by the Microelectronics Advanced Research Corp. (MARCO) and the Defense Advanced Research Projects Agency (DARPA).

Here’s a link to and a citation for the paper,

Unidirectional spin-Hall and Rashba−Edelstein magnetoresistance in topological insulator-ferromagnet layer heterostructures by Yang Lv, James Kally, Delin Zhang, Joon Sue Lee, Mahdi Jamali, Nitin Samarth, & Jian-Ping Wang. Nature Communications 9, Article number: 111 (2018) doi:10.1038/s41467-017-02491-3 Published online: 09 January 2018

This is an open access paper.

From the memristor to the atomristor?

I’m going to let Michael Berger explain the memristor (from Berger’s Jan. 2, 2017 Nanowerk Spotlight article),

In trying to bring brain-like (neuromorphic) computing closer to reality, researchers have been working on the development of memory resistors, or memristors, which are resistors in a circuit that ‘remember’ their state even if you lose power.

Today, most computers use random access memory (RAM), which moves very quickly as a user works but does not retain unsaved data if power is lost. Flash drives, on the other hand, store information when they are not powered but work much slower. Memristors could provide a memory that is the best of both worlds: fast and reliable.

He goes on to discuss a team at the University of Texas at Austin’s work on creating an extraordinarily thin memristor: an atomristor,

he team’s work features the thinnest memory devices and it appears to be a universal effect available in all semiconducting 2D monolayers.

The scientists explain that the unexpected discovery of nonvolatile resistance switching (NVRS) in monolayer transitional metal dichalcogenides (MoS2, MoSe2, WS2, WSe2) is likely due to the inherent layered crystalline nature that produces sharp interfaces and clean tunnel barriers. This prevents excessive leakage and affords stable phenomenon so that NVRS can be used for existing memory and computing applications.

“Our work opens up a new field of research in exploiting defects at the atomic scale, and can advance existing applications such as future generation high density storage, and 3D cross-bar networks for neuromorphic memory computing,” notes Akinwande [Deji Akinwande, an Associate Professor at the University of Texas at Austin]. “We also discovered a completely new application, which is non-volatile switching for radio-frequency (RF) communication systems. This is a rapidly emerging field because of the massive growth in wireless technologies and the need for very low-power switches. Our devices consume no static power, an important feature for battery life in mobile communication systems.”

Here’s a link to and a citation for the Akinwande team’s paper,

Atomristor: Nonvolatile Resistance Switching in Atomic Sheets of Transition Metal Dichalcogenides by Ruijing Ge, Xiaohan Wu, Myungsoo Kim, Jianping Shi, Sushant Sonde, Li Tao, Yanfeng Zhang, Jack C. Lee, and Deji Akinwande. Nano Lett., Article ASAP DOI: 10.1021/acs.nanolett.7b04342 Publication Date (Web): December 13, 2017

Copyright © 2017 American Chemical Society

This paper appears to be open access.

ETA January 23, 2018: There’s another account of the atomristor in Samuel K. Moore’s January 23, 2018 posting on the Nanoclast blog (on the IEEE [Institute of Electrical and Electronics Engineers] website).

Leftover 2017 memristor news bits

i have two bits of news, one from this October 2017 about using light to control a memristor’s learning properties and one from December 2017 about memristors and neural networks.

Shining a light on the memristor

Michael Berger wrote an October 30, 2017 Nanowerk Sportlight article about some of the latest work concerning memristors and light,

Memristors – or resistive memory – are nanoelectronic devices that are very promising components for next generation memory and computing devices. They are two-terminal electric elements similar to a conventional resistor – however, the electric resistance in a memristor is dependent on the charge passing through it; which means that its conductance can be precisely modulated by charge or flux through it. Its special property is that its resistance can be programmed (resistor function) and subsequently remains stored (memory function).

In this sense, a memristor is similar to a synapse in the human brain because it exhibits the same switching characteristics, i.e. it is able, with a high level of plasticity, to modify the efficiency of signal transfer between neurons under the influence of the transfer itself. That’s why researchers are hopeful to use memristors for the fabrication of electronic synapses for neuromorphic (i.e. brain-like) computing that mimics some of the aspects of learning and computation in human brains.

Human brains may be slow at pure number crunching but they are excellent at handling fast dynamic sensory information such as image and voice recognition. Walking is something that we take for granted but this is quite challenging for robots, especially over uneven terrain.

“Memristors present an opportunity to make new types of computers that are different from existing von Neumann architectures, which traditional computers are based upon,” Dr Neil T. Kemp, a Lecturer in Physics at the University of Hull [UK], tells Nanowerk. “Our team at the University of Hull is focussed on making memristor devices dynamically reconfigurable and adaptive – we believe this is the route to making a new generation of artificial intelligence systems that are smarter and can exhibit complex behavior. Such systems would also have the advantage of memristors, high density integration and lower power usage, so these systems would be more lightweight, portable and not need re-charging so often – which is something really needed for robots etc.”

In their new paper in Nanoscale (“Reversible Optical Switching Memristors with Tunable STDP Synaptic Plasticity: A Route to Hierarchical Control in Artificial Intelligent Systems”), Kemp and his team demonstrate the ability to reversibly control the learning properties of memristors via optical means.

The reversibility is achieved by changing the polarization of light. The researchers have used this effect to demonstrate tuneable learning in a memristor. One way this is achieved is through something called Spike Timing Dependent Plasticity (STDP), which is an effect known to occur in human brains and is linked with sensory perception, spatial reasoning, language and conscious thought in the neocortex.

STDP learning is based upon differences in the arrival time of signals from two adjacent neurons. The University of Hull team has shown that they can modulate the synaptic plasticity via optical means which enables the devices to have tuneable learning.

“Our research findings are important because it demonstrates that light can be used to control the learning properties of a memristor,” Kemp points out. “We have shown that light can be used in a reversible manner to change the connection strength (or conductivity) of artificial memristor synapses and as well control their ability to forget i.e. we can dynamically change device to have short-term or long-term memory.”

According to the team, there are many potential applications, such as adaptive electronic circuits controllable via light, or in more complex systems, such as neuromorphic computing, the development of optically reconfigurable neural networks.

Having optically controllable memristors can also facilitate the implementation of hierarchical control in larger artificial-brain like systems, whereby some of the key processes that are carried out by biological molecules in human brains can be emulated in solid-state devices through patterning with light.

Some of these processes include synaptic pruning, conversion of short term memory to long term memory, erasing of certain memories that are no longer needed or changing the sensitivity of synapses to be more adept at learning new information.

“The ability to control this dynamically, both spatially and temporally, is particularly interesting since it would allow neural networks to be reconfigurable on the fly through either spatial patterning or by adjusting the intensity of the light source,” notes Kemp.

In their new paper in Nanoscale Currently, the devices are more suited to neuromorphic computing applications, which do not need to be as fast. Optical control of memristors opens the route to dynamically tuneable and reprogrammable synaptic circuits as well the ability (via optical patterning) to have hierarchical control in larger and more complex artificial intelligent systems.

“Artificial Intelligence is really starting to come on strong in many areas, especially in the areas of voice/image recognition and autonomous systems – we could even say that this is the next revolution, similarly to what the industrial revolution was to farming and production processes,” concludes Kemp. “There are many challenges to overcome though. …

That excerpt should give you the gist of Berger’s article and, for those who need more information, there’s Berger’s article and, also, a link to and a citation for the paper,

Reversible optical switching memristors with tunable STDP synaptic plasticity: a route to hierarchical control in artificial intelligent systems by Ayoub H. Jaafar, Robert J. Gray, Emanuele Verrelli, Mary O’Neill, Stephen. M. Kelly, and Neil T. Kemp. Nanoscale, 2017,9, 17091-17098 DOI: 10.1039/C7NR06138B First published on 24 Oct 2017

This paper is behind a paywall.

The memristor and the neural network

It would seem machine learning could experience a significant upgrade if the work in Wei Lu’s University of Michigan laboratory can be scaled for general use. From a December 22, 2017 news item on ScienceDaily,

A new type of neural network made with memristors can dramatically improve the efficiency of teaching machines to think like humans.

The network, called a reservoir computing system, could predict words before they are said during conversation, and help predict future outcomes based on the present.

The research team that created the reservoir computing system, led by Wei Lu, professor of electrical engineering and computer science at the University of Michigan, recently published their work in Nature Communications.

A December 19, 2017 University of Michigan news release (also on EurekAlert) by Dan Newman, which originated the news item, expands on the theme,

Reservoir computing systems, which improve on a typical neural network’s capacity and reduce the required training time, have been created in the past with larger optical components. However, the U-M group created their system using memristors, which require less space and can be integrated more easily into existing silicon-based electronics.

Memristors are a special type of resistive device that can both perform logic and store data. This contrasts with typical computer systems, where processors perform logic separate from memory modules. In this study, Lu’s team used a special memristor that memorizes events only in the near history.

Inspired by brains, neural networks are composed of neurons, or nodes, and synapses, the connections between nodes.

To train a neural network for a task, a neural network takes in a large set of questions and the answers to those questions. In this process of what’s called supervised learning, the connections between nodes are weighted more heavily or lightly to minimize the amount of error in achieving the correct answer.

Once trained, a neural network can then be tested without knowing the answer. For example, a system can process a new photo and correctly identify a human face, because it has learned the features of human faces from other photos in its training set.

“A lot of times, it takes days or months to train a network,” says Lu. “It is very expensive.”

Image recognition is also a relatively simple problem, as it doesn’t require any information apart from a static image. More complex tasks, such as speech recognition, can depend highly on context and require neural networks to have knowledge of what has just occurred, or what has just been said.

“When transcribing speech to text or translating languages, a word’s meaning and even pronunciation will differ depending on the previous syllables,” says Lu.

This requires a recurrent neural network, which incorporates loops within the network that give the network a memory effect. However, training these recurrent neural networks is especially expensive, Lu says.

Reservoir computing systems built with memristors, however, can skip most of the expensive training process and still provide the network the capability to remember. This is because the most critical component of the system – the reservoir – does not require training.

When a set of data is inputted into the reservoir, the reservoir identifies important time-related features of the data, and hands it off in a simpler format to a second network. This second network then only needs training like simpler neural networks, changing weights of the features and outputs that the first network passed on until it achieves an acceptable level of error.

Enlargereservoir computing system

IMAGE:  Schematic of a reservoir computing system, showing the reservoir with internal dynamics and the simpler output. Only the simpler output needs to be trained, allowing for quicker and lower-cost training. Courtesy Wei Lu.

 

“The beauty of reservoir computing is that while we design it, we don’t have to train it,” says Lu.

The team proved the reservoir computing concept using a test of handwriting recognition, a common benchmark among neural networks. Numerals were broken up into rows of pixels, and fed into the computer with voltages like Morse code, with zero volts for a dark pixel and a little over one volt for a white pixel.

Using only 88 memristors as nodes to identify handwritten versions of numerals, compared to a conventional network that would require thousands of nodes for the task, the reservoir achieved 91% accuracy.

Reservoir computing systems are especially adept at handling data that varies with time, like a stream of data or words, or a function depending on past results.

To demonstrate this, the team tested a complex function that depended on multiple past results, which is common in engineering fields. The reservoir computing system was able to model the complex function with minimal error.

Lu plans on exploring two future paths with this research: speech recognition and predictive analysis.

“We can make predictions on natural spoken language, so you don’t even have to say the full word,” explains Lu.

“We could actually predict what you plan to say next.”

In predictive analysis, Lu hopes to use the system to take in signals with noise, like static from far-off radio stations, and produce a cleaner stream of data. “It could also predict and generate an output signal even if the input stopped,” he says.

EnlargeWei Lu

IMAGE:  Wei Lu, Professor of Electrical Engineering & Computer Science at the University of Michigan holds a memristor he created. Photo: Marcin Szczepanski.

 

The work was published in Nature Communications in the article, “Reservoir computing using dynamic memristors for temporal information processing”, with authors Chao Du, Fuxi Cai, Mohammed Zidan, Wen Ma, Seung Hwan Lee, and Prof. Wei Lu.

The research is part of a $6.9 million DARPA [US Defense Advanced Research Projects Agency] project, called “Sparse Adaptive Local Learning for Sensing and Analytics [also known as SALLSA],” that aims to build a computer chip based on self-organizing, adaptive neural networks. The memristor networks are fabricated at Michigan’s Lurie Nanofabrication Facility.

Lu and his team previously used memristors in implementing “sparse coding,” which used a 32-by-32 array of memristors to efficiently analyze and recreate images.

Here’s a link to and a citation for the paper,

Reservoir computing using dynamic memristors for temporal information processing by Chao Du, Fuxi Cai, Mohammed A. Zidan, Wen Ma, Seung Hwan Lee & Wei D. Lu. Nature Communications 8, Article number: 2204 (2017) doi:10.1038/s41467-017-02337-y Published online: 19 December 2017

This is an open access paper.

Mott memristor

Mott memristors (mentioned in my Aug. 24, 2017 posting about neuristors and brainlike computing) gets more fulsome treatment in an Oct. 9, 2017 posting by Samuel K. Moore on the Nanoclast blog (found on the IEEE [Institute of Electrical and Electronics Engineers] website) Note: 1: Links have been removed; Note 2 : I quite like Moore’s writing style but he’s not for the impatient reader,

When you’re really harried, you probably feel like your head is brimful of chaos. You’re pretty close. Neuroscientists say your brain operates in a regime termed the “edge of chaos,” and it’s actually a good thing. It’s a state that allows for fast, efficient analog computation of the kind that can solve problems that grow vastly more difficult as they become bigger in size.

The trouble is, if you’re trying to replicate that kind of chaotic computation with electronics, you need an element that both acts chaotically—how and when you want it to—and could scale up to form a big system.

“No one had been able to show chaotic dynamics in a single scalable electronic device,” says Suhas Kumar, a researcher at Hewlett Packard Labs, in Palo Alto, Calif. Until now, that is.

He, John Paul Strachan, and R. Stanley Williams recently reported in the journal Nature that a particular configuration of a certain type of memristor contains that seed of controlled chaos. What’s more, when they simulated wiring these up into a type of circuit called a Hopfield neural network, the circuit was capable of solving a ridiculously difficult problem—1,000 instances of the traveling salesman problem—at a rate of 10 trillion operations per second per watt.

(It’s not an apples-to-apples comparison, but the world’s most powerful supercomputer as of June 2017 managed 93,015 trillion floating point operations per second but consumed 15 megawatts doing it. So about 6 billion operations per second per watt.)

The device in question is called a Mott memristor. Memristors generally are devices that hold a memory, in the form of resistance, of the current that has flowed through them. The most familiar type is called resistive RAM (or ReRAM or RRAM, depending on who’s asking). Mott memristors have an added ability in that they can also reflect a temperature-driven change in resistance.

The HP Labs team made their memristor from an 8-nanometer-thick layer of niobium dioxide (NbO2) sandwiched between two layers of titanium nitride. The bottom titanium nitride layer was in the form of a 70-nanometer wide pillar. “We showed that this type of memristor can generate chaotic and nonchaotic signals,” says Williams, who invented the memristor based on theory by Leon Chua.

(The traveling salesman problem is one of these. In it, the salesman must find the shortest route that lets him visit all of his customers’ cities, without going through any of them twice. It’s a difficult problem because it becomes exponentially more difficult to solve with each city you add.)

Here’s what the niobium dioxide-based Mott memristor looks like,

Photo: Suhas Kumar/Hewlett Packard Labs
A micrograph shows the construction of a Mott memristor composed of an 8-nanometer-thick layer of niobium dioxide between two layers of titanium nitride.

Here’s a link to and a citation for the paper,

Chaotic dynamics in nanoscale NbO2 Mott memristors for analogue computing by Suhas Kumar, John Paul Strachan & R. Stanley Williams. Nature 548, 318–321 (17 August 2017) doi:10.1038/nature23307 Published online: 09 August 2017

This paper is behind a paywall.

Machine learning software and quantum computers that think

A Sept. 14, 2017 news item on phys.org sets the stage for quantum machine learning by explaining a few basics first,

Language acquisition in young children is apparently connected with their ability to detect patterns. In their learning process, they search for patterns in the data set that help them identify and optimize grammar structures in order to properly acquire the language. Likewise, online translators use algorithms through machine learning techniques to optimize their translation engines to produce well-rounded and understandable outcomes. Even though many translations did not make much sense at all at the beginning, in these past years we have been able to see major improvements thanks to machine learning.

Machine learning techniques use mathematical algorithms and tools to search for patterns in data. These techniques have become powerful tools for many different applications, which can range from biomedical uses such as in cancer reconnaissance, in genetics and genomics, in autism monitoring and diagnosis and even plastic surgery, to pure applied physics, for studying the nature of materials, matter or even complex quantum systems.

Capable of adapting and changing when exposed to a new set of data, machine learning can identify patterns, often outperforming humans in accuracy. Although machine learning is a powerful tool, certain application domains remain out of reach due to complexity or other aspects that rule out the use of the predictions that learning algorithms provide.

Thus, in recent years, quantum machine learning has become a matter of interest because of is vast potential as a possible solution to these unresolvable challenges and quantum computers show to be the right tool for its solution.

A Sept. 14, 2017 Institute of Photonic Sciences ([Catalan] Institut de Ciències Fotòniques] ICFO) press release, which originated the news item, goes on to detail a recently published overview of the state of quantum machine learning,

In a recent study, published in Nature, an international team of researchers integrated by Jacob Biamonte from Skoltech/IQC, Peter Wittek from ICFO, Nicola Pancotti from MPQ, Patrick Rebentrost from MIT, Nathan Wiebe from Microsoft Research, and Seth Lloyd from MIT have reviewed the actual status of classical machine learning and quantum machine learning. In their review, they have thoroughly addressed different scenarios dealing with classical and quantum machine learning. In their study, they have considered different possible combinations: the conventional method of using classical machine learning to analyse classical data, using quantum machine learning to analyse both classical and quantum data, and finally, using classical machine learning to analyse quantum data.

Firstly, they set out to give an in-depth view of the status of current supervised and unsupervised learning protocols in classical machine learning by stating all applied methods. They introduce quantum machine learning and provide an extensive approach on how this technique could be used to analyse both classical and quantum data, emphasizing that quantum machines could accelerate processing timescales thanks to the use of quantum annealers and universal quantum computers. Quantum annealing technology has better scalability, but more limited use cases. For instance, the latest iteration of D-Wave’s [emphasis mine] superconducting chip integrates two thousand qubits, and it is used for solving certain hard optimization problems and for efficient sampling. On the other hand, universal (also called gate-based) quantum computers are harder to scale up, but they are able to perform arbitrary unitary operations on qubits by sequences of quantum logic gates. This resembles how digital computers can perform arbitrary logical operations on classical bits.

However, they address the fact that controlling a quantum system is very complex and analyzing classical data with quantum resources is not as straightforward as one may think, mainly due to the challenge of building quantum interface devices that allow classical information to be encoded into a quantum mechanical form. Difficulties, such as the “input” or “output” problems appear to be the major technical challenge that needs to be overcome.

The ultimate goal is to find the most optimized method that is able to read, comprehend and obtain the best outcomes of a data set, be it classical or quantum. Quantum machine learning is definitely aimed at revolutionizing the field of computer sciences, not only because it will be able to control quantum computers, speed up the information processing rates far beyond current classical velocities, but also because it is capable of carrying out innovative functions, such quantum deep learning, that could not only recognize counter-intuitive patterns in data, invisible to both classical machine learning and to the human eye, but also reproduce them.

As Peter Wittek [emphasis mine] finally states, “Writing this paper was quite a challenge: we had a committee of six co-authors with different ideas about what the field is, where it is now, and where it is going. We rewrote the paper from scratch three times. The final version could not have been completed without the dedication of our editor, to whom we are indebted.”

It was a bit of a surprise to see local (Vancouver, Canada) company D-Wave Systems mentioned but i notice that one of the paper’s authors (Peter Wittek) is mentioned in a May 22, 2017 D-Wave news release announcing a new partnership to foster quantum machine learning,

Today [May 22, 2017] D-Wave Systems Inc., the leader in quantum computing systems and software, announced a new initiative with the Creative Destruction Lab (CDL) at the University of Toronto’s Rotman School of Management. D-Wave will work with CDL, as a CDL Partner, to create a new track to foster startups focused on quantum machine learning. The new track will complement CDL’s successful existing track in machine learning. Applicants selected for the intensive one-year program will go through an introductory boot camp led by Dr. Peter Wittek [emphasis mine], author of Quantum Machine Learning: What Quantum Computing means to Data Mining, with instruction and technical support from D-Wave experts, access to a D-Wave 2000Q™ quantum computer, and the opportunity to use a D-Wave sampling service to enable machine learning computations and applications. D-Wave staff will be a part of the committee selecting up to 40 individuals for the program, which begins in September 2017.

For anyone interested in the paper, here’s a link to and a citation,

Quantum machine learning by Jacob Biamonte, Peter Wittek, Nicola Pancotti, Patrick Rebentrost, Nathan Wiebe, & Seth Lloyd. Nature 549, 195–202 (14 September 2017) doi:10.1038/nature23474 Published online 13 September 2017

This paper is behind a paywall.

Brain composer

This is a representation of the work they are doing on brain-computer interfaces (BCI) at the Technical University of Graz (TU Graz; Austria),

A Sept. 11, 2017 news item on phys.org announces the research into thinking melodies turning them into a musical score,

TU Graz researchers develop new brain-computer interface application that allows music to be composed by the power of thought. They have published their results in the current issue of the journal PLOS ONE.

Brain-computer interfaces (BCI) can replace bodily functions to a certain degree. Thanks to BCI, physically impaired persons can control special prostheses via their minds, surf the internet and write emails.

A group led by BCI expert Gernot Müller-Putz from TU Graz’s Institute of Neural Engineering shows that experiences of quite a different tone can be sounded from the keys of brain-computer interfaces. Derived from an established BCI method for writing, the team has developed a new application by which music can be composed and transferred onto a musical score through the power of thought. It employs a special cap that measures brain waves, the adapted BCI, music composition software, and a bit of musical knowledge.

A Sept. 6, 2017 TU Graz press release by Suzanne Eigner, which originated the news item, explains the research in more detail,

The basic principle of the BCI method used, which is called P300, can be briefly described: various options, such as letters or notes, pauses, chords, etc. flash by one after the other in a table. If you’re trained and can focus on the desired option while it lights up, you cause a minute change in your brain waves. The BCI recognises this change and draws conclusions about the chosen option.

Musical test persons

18 test persons chosen for the study by Gernot Müller-Putz, Andreas Pinegger and Selina C. Wriessnegger from TU Graz’s Institute of Neural Engineering as well as Hannah Hiebel, meanwhile at the Institute of Cognitive Psychology & Neuroscience at the University of Graz, had to “think” melodies onto a musical score. All test subjects were of sound bodily health during the study and had a certain degree of basic musical and compositional knowledge since they all played musical instruments to some degree. Among the test persons was the late Graz composer and clarinettist, Franz Cibulka. “The results of the BCI compositions can really be heard. And what is more important: the test persons enjoyed it. After a short training session, all of them could start composing and seeing their melodies on the score and then play them. The very positive results of the study with bodily healthy test persons are the first step in a possible expansion of the BCI composition to patients,” stresses Müller-Putz.

Sideshow of BCI research

This little-noticed sideshow of the lively BCI research at TU Graz, with its distinct focus on disabled persons, shows us which other avenues may yet be worth exploring. Meanwhile there are some initial attempts at BCI systems on smart phones. This makes it easier for people to use BCI applications, since the smart phone as powerful computer is becoming part of the BCI system. It is thus conceivable, for instance, to have BCI apps which can analyse brain signals for various applications. “20 years ago, the idea of composing a piece of music using the power of the mind was unimaginable. Now we can do it, and at the same time have tens of new, different ideas which are in part, once again, a long way from becoming reality. We still need a bit more time before it is mature enough for daily applications. The BCI community is working in many directions at high pressure.

Here’s a link to and a citation for the paper,

Composing only by thought: Novel application of the P300 brain-computer interface by Andreas Pinegger, Hannah Hiebel, Selina C. Wriessnegger, Gernot R. Müller-Putz. PLOS https://doi.org/10.1371/journal.pone.0181584 Published: September 6, 2017

This paper is open access.

This BCI ‘sideshow’ reminded me of The Music Man, a musical by Meredith Wilson. It was both a play and a film  and I’ve only ever seen the 1962 film. It features a con man, Harold Hill, who sells musical instruments and uniforms in small towns in Iowa. He has no musical training but while he’s conning the townspeople he convinces them that he can provide musical training with his ‘think method’. After falling in love with one of the townsfolk, he is hunted down and made to prove his method works. This is a clip from a Broadway revival of the play where Harold Hill is hoping that his ‘think method’ while yield results,

Of course, the people in this study had musicaltraining so they could think a melody into a musical score but I find the echo from the past amusing nonetheless.