Computers Math

The team, led by Bristol researcher and Phasecraft co-founder, Dr. Ashley Montanaro, has discovered algorithms and analysis which significantly lessen the quantum hardware capability needed to solve problems which go beyond the realm of classical computing, even supercomputers.
In the paper, published in Physical Review B, the team demonstrates how optimised quantum algorithms can solve the notorious Fermi-Hubbard model on near-term hardware.
The Fermi-Hubbard model is of fundamental importance in condensed-matter physics as a model for strongly correlated materials and a route to understanding high-temperature superconductivity.
Finding the ground state of the Fermi-Hubbard model has been predicted to be one of the first applications of near-term quantum computers, and one that offers a pathway to understanding and developing novel materials.
Dr. Ashley Montanaro, research lead and cofounder of Phasecraft: “Quantum computing has critically important applications in materials science and other domains. Despite the major quantum hardware advances recently, we may still be several years from having the right software and hardware to solve meaningful problems with quantum computing. Our research focuses on algorithms and software optimisations to maximise the quantum hardware’s capacity, and bring quantum computing closer to reality.
“Near-term quantum hardware will have limited device and computation size. Phasecraft applied new theoretical ideas and numerical experiments to put together a very comprehensive study on different strategies for solving the Fermi-Hubbard model, zeroing in on strategies that are most likely to have the best results and impact in the near future.

advertisement

“The results suggest that optimising over quantum circuits with a gate depth substantially less than a thousand could be sufficient to solve instances of the Fermi-Hubbard model beyond the capacity of a supercomputer. This new research shows significant promise for the capabilities of near-term quantum devices, improving on previous research findings by around a factor of 10.”
Physical Review B, published by the American Physical Society, is the top specialist journal in condensed-matter physics. The peer-reviewed research paper was also chosen as the Editors’ Suggestion and to appear in Physics magazine.
Andrew Childs, Professor in the Department of Computer Science and Institute for Advanced Computer Studies at the University of Maryland: “The Fermi-Hubbard model is a major challenge in condensed-matter physics, and the Phasecraft team has made impressive steps in showing how quantum computers could solve it. Their work suggests that surprisingly low-depth circuits could provide useful information about this model, making it more accessible to realistic quantum hardware.”
Hartmut Neven, Head of Quantum Artificial Intelligence Lab, Google: “Sooner or later, quantum computing is coming. Developing the algorithms and technology to power the first commercial applications of early quantum computing hardware is the toughest challenge facing the field, which few are willing to take on. We are proud to be partners with Phasecraft, a team that are developing advances in quantum software that could shorten that timeframe by years.”
Phasecraft Founder Dr. Toby Cubitt: “At Phasecraft, our team of leading quantum theorists have been researching and applying quantum theory for decades, leading some of the top global academic teams and research in the field. Today, Ashley and his team have demonstrated ways to get closer to achieving new possibilities that exist just beyond today’s technological bounds.”
Phasecraft has closed a record seed round for a quantum company in the UK with £3.7m in funding from private-sector VC investors, led by LocalGlobe with Episode1 along with previous investors. Former Songkick founder Ian Hogarth has also joined as board chair for Phasecraft. Phasecraft previously raised a £750,000 pre-seed round led by UCL Technology Fund with Parkwalk Advisors and London Co-investment Fund and has earned several grants facilitated by InnovateUK. Between equity funding and research grants, Phasecraft has raised more than £5.5m.
Dr Toby Cubitt: “With new funding and support, we are able to continue our pioneering research and industry collaborations to develop the quantum computing industry and find useful applications faster.”

Story Source:
Materials provided by University of Bristol. Note: Content may be edited for style and length. More

Researchers at Pohang University of Science and Technology (POSTECH) and Seoul National University in South Korea have demonstrated a new way to enhance the energy efficiency of a non-volatile magnetic memory device called SOT-MRAM. Published in Advanced Materials, this finding opens up a new window of exciting opportunities for future energy-efficient magnetic memories based on spintronics.
In modern computers, the random access memory (RAM) is used to store information. The SOT-MRAM (spin-orbit torque magnetic RAM) is one of the leading candidates for the next-generation memory technologies that aim to surpass the performance of various existing RAMs. The SOT-MRAM may operate faster than the fastest existing RAM (SRAM) and maintain information even after the electric energy supply is powered off whereas all fast RAMs existing today lose information as soon as the energy supply is powered off. The present level of the SOT-MRAM technology falls short of being satisfactory, however, due to its high energy demand; it requires large energy supply (or large current) to write information. Lowering the energy demand and enhancing the energy efficiency is an outstanding problem for the SOT-MRAM.
In the SOT-MRAM, magnetization directions of tiny magnets store information and writing amounts to change the magnetization directions to desired directions. The magnetization direction change is achieved by a special physics phenomenon called SOT that modifies the magnetization direction when a current is applied. To enhance the energy efficiency, soft magnets are ideal material choice for the tiny magnets since their magnetization directions can be easily alterned by a small current. Soft magnets are bad choice for the safe storage of information since their magnetization direction may be altered even when not intended — due to thermal noise or other noise. For this reason, most attempts to build the SOT-MRAM adopt hard magnets, because they magnetize very strongly and their magnetization direction is not easily altered by noise. But this material choice inevitably makes the energy efficiency of the SOT-MRAM poor.
A joint research team led by Professor Hyun-Woo Lee in the Department of Physics at POSTECH and Professor Je-Geun Park in the Department of Physics at Seoul National University (former associate director of the Center for Correlated Electron Systems within the Institute for Basic Science in Korea), demonstrated a way to enhance the energy efficiency without sacrificing the demand for safe storage. They reported that ultrathin iron germanium telluride (Fe3GeTe2, FGT) — a ferromagnetic material with special geometrical symmetry and quantum properties — switches from a hard magnet to a soft magnet when a small current is applied. Thus when information writing is not intended, the material remains a hard magnet, which is good for the safe storage, and only when writing is intended, the material switches to a soft magnet, allowing for enhanced energy efficiency.
“Intriguing properties of layered materials never cease to amaze me: the current through FGT induces a highly unusual type of spin-orbit torque (SOT), which modifies the energy profile of this material to switch it from a hard magnet to a soft magnet. This is in clear contrast to SOT produced by other materials, which may change the magnetization direction but cannot switch a hard magnet to a soft magnet,” explains Professor Lee.
Experiments by Professor Park’s group revealed that this FGT-based magnetic memory device is highly energy-efficient. In particular, the measured magnitude of SOT per applied current density is two orders of magnitude larger than the values reported previously for other candidate materials for the SOT-MRAM.
“Controlling magnetic states with a small current is essential for the next-generation of energy-efficient devices. These will be able to store greater amounts of data and enable faster data access than today’s electronic memories, while consuming less energy,” notes Dr. Kaixuan Zhang who is a team leader in Professor Park’s group, interested in studying the application of correlated quantum physics in spintronic devices.
“Our findings open up a fascinating avenue of electrical modulation and spintronic applications using 2D layered magnetic materials,” closed Professor Lee.

Story Source:
Materials provided by Pohang University of Science & Technology (POSTECH). Note: Content may be edited for style and length. More

Researchers at Hokkaido University and Amoeba Energy in Japan have, inspired by the efficient foraging behavior of a single-celled amoeba, developed an analog computer for finding a reliable and swift solution to the traveling salesman problem — a representative combinatorial optimization problem.
Many real-world application tasks such as planning and scheduling in logistics and automation are mathematically formulated as combinatorial optimization problems. Conventional digital computers, including supercomputers, are inadequate to solve these complex problems in practically permissible time as the number of candidate solutions they need to evaluate increases exponentially with the problem size — also known as combinatorial explosion. Thus new computers called “Ising machines,” including “quantum annealers,” have been actively developed in recent years. These machines, however, require complicated pre-processing to convert each task to the form they can handle and have a risk of presenting illegal solutions that do not meet some constraints and requests, resulting in major obstacles to the practical applications.
These obstacles can be avoided using the newly developed “electronic amoeba,” an analog computer inspired by a single-celled amoeboid organism. The amoeba is known to maximize nutrient acquisition efficiently by deforming its body. It has shown to find an approximate solution to the traveling salesman problem (TSP), i.e., given a map of a certain number of cities, the problem is to find the shortest route for visiting each city exactly once and returning to the starting city. This finding inspired Professor Seiya Kasai at Hokkaido University to mimic the dynamics of the amoeba electronically using an analog circuit, as described in the journal Scientific Reports. “The amoeba core searches for a solution under the electronic environment where resistance values at intersections of crossbars represent constraints and requests of the TSP,” says Kasai. Using the crossbars, the city layout can be easily altered by updating the resistance values without complicated pre-processing.
Kenta Saito, a PhD student in Kasai’s lab, fabricated the circuit on a breadboard and succeeded in finding the shortest route for the 4-city TSP. He evaluated the performance for larger-sized problems using a circuit simulator. Then the circuit reliably found a high-quality legal solution with a significantly shorter route length than the average length obtained by the random sampling. Moreover, the time required to find a high-quality legal solution grew only linearly to the numbers of cities. Comparing the search time with a representative TSP algorithm “2-opt,” the electronic amoeba becomes more advantageous as the number of cities increases. “The analog circuit reproduces well the unique and efficient optimization capability of the amoeba, which the organism has acquired through natural selection,” says Kasai.
“As the analog computer consists of a simple and compact circuit, it can tackle many real-world problems in which inputs, constraints, and requests dynamically change and can be embedded into IoT devices as a power-saving microchip,” says Masashi Aono who leads Amoeba Energy to promote the practical use of the amoeba-inspired computers.
This is a Joint Release between Hokkaido University and Amoeba Energy Co., Ltd.

Story Source:
Materials provided by Hokkaido University. Note: Content may be edited for style and length. More

Robotic animals could be the ‘pawfect’ replacement for our real-life furry friends, a new study published today by the University of Portsmouth has found.
Animals, especially dogs, can have therapeutic benefits for children and young people. A new paper, published in The International Journal of Social Robotics, has found that the robotic animal, ‘MiRo-E’, can be just as effective and may even be a better alternative.
Dr Leanne Proops from the Department of Psychology, who supervised the study said: “We know that real dogs can provide calming and enjoyable interactions for children — increasing their feelings of wellbeing, improving motivation and reducing stress.
“This preliminary study has found that biomimetic robots — robots that mimic animal behaviours — may be a suitable replacement in certain situations and there are some benefits to using them over a real dog.”
Dogs are the most commonly used animals for therapy because of their training potential and generally social nature. However, there are concerns about using them in a setting with children because of the risk of triggering allergies or transmitting disease, and some people do not like dogs, so may not be comfortable in the presence of a real therapy dog.
Olivia Barber, who owns a therapy dog herself, and is first author of the paper, said: “Although lots of people in schools and hospitals benefit greatly from receiving visits from a therapy dog, we have to be mindful of the welfare of the therapy dog. Visits can be stressful and incredibly tiring for therapy dogs, meaning that we should be exploring whether using a robotic animal is feasible.”
There are lots of positives to using a robotic animal over a therapy dog. They can be thoroughly cleaned and can work for longer periods of time. They can also be incredibly lifelike, mirroring the movements and behaviour of a real animal, such as wagging their tails to show excitement, expressing “emotions” through sounds and colour, turning their ears towards sounds and even going to sleep.

advertisement

The researchers used real dogs and a biomimetic robot in a mainstream secondary school in West Sussex to interact with 34 children aged 11-12.
The two real-life therapy dogs were a three-year-old Jack Russell crossed with a Poodle and a 12-year-old Labrador-retriever from the charity Pets as Therapy. The robot was a MiRo-E biomimetic robot developed by Consequential Robotics.
The children were asked to complete a questionnaire about their beliefs and attitudes towards dogs and robots, before they took part in two separate free-play sessions, one with a real-life dog and one with a robot.
The researchers found the children spent a similar amount of time stroking both the real-life dog and the robot, but they spent more time interacting with the robot.
Despite the children reporting they significantly preferred the session with the living dog, overall enjoyment was high and they actually expressed more positive emotions following interaction with the robot. The more the children attributed mental states and sentience to the dog and robot, the more they enjoyed the sessions.
Dr Proops said: “This is a small-scale study, but the results show that interactive robotic animals could be used as a good comparison to live dogs in research, and a useful alternative to traditional animal therapy.”

Story Source:
Materials provided by University of Portsmouth. Note: Content may be edited for style and length. More

Although robotics has reshaped and even redefined many industrial sectors, there still exists a gap between machines and humans in fields such as health and elderly care. For robots to safely manipulate or interact with fragile objects and living organisms, new strategies to enhance their perception while making their parts softer are needed. In fact, building a safe and dexterous robotic gripper with human-like capabilities is currently one of the most important goals in robotics.
One of the main challenges in the design of soft robotic grippers is integrating traditional sensors onto the robot’s fingers. Ideally, a soft gripper should have what’s known as proprioception — a sense of its own movements and position — to be able to safely execute varied tasks. However, traditional sensors are rigid and compromise the mechanical characteristics of the soft parts. Moreover, existing soft grippers are usually designed with a single type of proprioceptive sensation; either pressure or finger curvature.
To overcome these limitations, scientists at Ritsumeikan University, Japan, have been working on novel soft gripper designs under the lead of Associate Professor Mengying Xie. In their latest study published in Nano Energy, they successfully used multimaterial 3D printing technology to fabricate soft robotic fingers with a built-in proprioception sensor. Their design strategy offers numerous advantages and represents a large step toward safer and more capable soft robots.
The soft finger has a reinforced inflation chamber that makes it bend in a highly controllable way according to the input air pressure. In addition, the stiffness of the finger is also tunable by creating a vacuum in a separate chamber. This was achieved through a mechanism called vacuum jamming, by which multiple stacked layers of a bendable material can be made rigid by sucking out the air between them. Both functions combined enable a three-finger robotic gripper to properly grasp and maintain hold of any object by ensuring the necessary force is applied.
Most notable, however, is that a single piezoelectric layer was included among the vacuum jamming layers as a sensor. The piezoelectric effect produces a voltage difference when the material is under pressure. The scientists leveraged this phenomenon as a sensing mechanism for the robotic finger, providing a simple way to sense both its curvature and initial stiffness (prior to vacuum adjustment). They further enhanced the finger’s sensitivity by including a microstructured layer among the jamming layers to improve the distribution of pressure on the piezoelectric material.
The use of multimaterial 3D printing, a simple and fast prototyping process, allowed the researchers to easily integrate the sensing and stiffness-tuning mechanisms into the design of the robotic finger itself. “Our work suggests a way of designing sensors that contribute not only as sensing elements for robotic applications, but also as active functional materials to provide better control of the whole system without compromising its dynamic behavior,” says Prof Xie. Another remarkable feature of their design is that the sensor is self-powered by the piezoelectric effect, meaning that it requires no energy supply — essential for low-power applications.
Overall, this exciting new study will help future researchers find new ways of improving how soft grippers interact with and sense the objects being manipulated. In turn, this will greatly expand the uses of robots, as Prof Xie indicates: “Self-powered built-in sensors will not only allow robots to safely interact with humans and their environment, but also eliminate the barriers to robotic applications that currently rely on powered sensors to monitor conditions.”
Let’s hope this technology is further developed so that our mechanical friends can soon join us in many more human activities!

Story Source:
Materials provided by Ritsumeikan University. Note: Content may be edited for style and length. More

Computer scientists have developed an entirely new way of interacting with video content that adapts to, and is controlled by, your body movement.
Fitness videos and other instructional content that aims to teach viewers new martial arts skills, exercises or yoga positions have been popular since VHS in the 80s and are abundant on Internet platforms like YouTube.
However, these traditional forms of instructional videos can lead to frustration, and even the potential for physical strain, as novice viewers, or those with limited physical mobility, struggle to keep up and mimic the movements of the expert instructors.
Now an international team of researchers from Lancaster University, Stanford University and FXPAL, have created a solution that dynamically adapts to mirror the position of the viewer’s body and matches the speed of video playback to the viewer’s movements.
The system, called ‘Reactive Video’, uses a Microsoft Kinect sensor, the latest in skeleton-tracking software, and probabilistic algorithms to identify the position, and movement of joints and limbs — such as elbows, knees, arms, hands, hips and legs. By working out the viewer’s movements it can match and compare this with the movement of the instructor in the video footage. It then estimates the time the user will take to perform a movement and adjusts playback of the video to the correct position, and pace, of the viewer.
As well as providing a more immersive experience, Reactive Video also helps users to more accurately mimic and learn new movements.

advertisement

The researchers tested the system on study participants performing tai chi and radio exercises — a form of callisthenics popular in Japan. The results from the study showed that both systems could adapt to the users’ movements.
Dr Christopher Clarke, researcher from Lancaster University and co-author on the paper, said: “Since the 1980s, and especially now with the Internet, videos have helped people stay active and have offered a cheaper, more convenient alternative to gym memberships and personal trainers. However, traditional video players do have limitations — they can’t provide feedback, or adapt the pace and intensity of the physical movement to the user.
“We know performing movements in slow motion is beneficial for learning by providing opportunities to analyse your movements, and developing timing. We also know it can result in less physical strain for inexperienced users.
“For some people, keeping pace can be tricky — especially when learning something new, and for older people or those with impaired movement. Also, constantly reaching for a remote to pause, rewind and replay, can be frustrating and breaks the immersion.
“Our system overcomes these issues by having the video automatically adjust itself to play back at the user’s speed, which is less stressful and more beneficial for learning.”
Don Kimber, co-author of the research, said: “Reactive Video acts and feels like a magic mirror where as you move the video mirrors your movement, but with a cleaned-up version of the procedure, or position, performed correctly by an expert for the user to mimic and learn from.”

advertisement

An additional benefit of Reactive Video, and something that sets it apart from exercise content developed for game consoles, is that it can be applied to existing footage of appropriate video content removing the need to create specially produced bespoke content.
“By using this system we can post-process existing instructional video content and enhance it to dynamically adapt to users providing a fundamental shift in how we can potentially interact with videos,” said Dr Clarke.
The team believe that with further research this kind of adaptive technology could be developed for sports and activities such as learning dance routines or honing golf swings.
The Reactive Video system was presented at UIST2020, a leading academic conference for the field of Human Computer Interaction.
It is detailed in the paper ‘Reactive Video: Adaptive Video Playback Based on User Motion for Supporting Physical Activity’.
The study’s authors are Christopher Clarke, of Lancaster University; Doga Cavdir of Stanford University; and Patrick Chiu, Laurent Denoue and Don Kimber, of FXPAL. More

For decades, one material has so dominated the production of computer chips and transistors that the tech capital of the world — Silicon Valley — bears its name. But silicon’s reign may not last forever.
MIT researchers have found that an alloy called InGaAs (indium gallium arsenide) could hold the potential for smaller and more energy efficient transistors. Previously, researchers thought that the performance of InGaAs transistors deteriorated at small scales. But the new study shows this apparent deterioration is not an intrinsic property of the material itself.
The finding could one day help push computing power and efficiency beyond what’s possible with silicon. “We’re really excited,” said Xiaowei Cai, the study’s lead author. “We hope this result will encourage the community to continue exploring the use of InGaAs as a channel material for transistors.”
Cai, now with Analog Devices, completed the research as a PhD student in the MIT Microsystems Technology Laboratories and Department of Electrical Engineering and Computer Science (EECS), with Donner Professor Jesús del Alamo. Her co-authors include Jesús Grajal of Polytechnic University of Madrid, as well as MIT’s Alon Vardi and del Alamo. The paper will be presented this month at the virtual IEEE International Electron Devices Meeting.
Transistors are the building blocks of a computer. Their role as switches, either halting electric current or letting it flow, gives rise to a staggering array of computations — from simulating the global climate to playing cat videos on Youtube. A single laptop could contain billions of transistors. For computing power to improve in the future, as it has for decades, electrical engineers will have to develop smaller, more tightly packed transistors. To date, silicon has been the semiconducting material of choice for transistors. But InGaAs has shown hints of becoming a potential competitor.
Electrons can zip through InGaAs with ease, even at low voltage. The material is “known to have great [electron] transport properties,” says Cai. InGaAs transistors can process signals quickly, potentially resulting in speedier calculations. Plus, InGaAs transistors can operate at relatively low voltage, meaning they could enhance a computer’s energy efficiency. So InGaAs might seem like a promising material for computer transistors. But there’s a catch.
InGaAs’ favorable electron transport properties seem to deteriorate at small scales — the scales needed to build faster and denser computer processors. The problem has led some researchers to conclude that nanoscale InGaAs transistors simply aren’t suited for the task. But, says Cai, “we have found that that’s a misconception.”
The team discovered that InGaAs’ small-scale performance issues are due in part to oxide trapping. This phenomenon causes electrons to get stuck while trying to flow through a transistor. “A transistor is supposed to work as a switch. You want to be able to turn a voltage on and have a lot of current,” says Cai. “But if you have electrons trapped, what happens is you turn a voltage on, but you only have a very limited amount of current in the channel. So the switching capability is a lot lower when you have that oxide trapping.”
Cai’s team pinpointed oxide trapping as the culprit by studying the transistor’s frequency dependence — the rate at which electric pulses are sent through the transistor. At low frequencies, the performance of nanoscale InGaAs transistors appeared degraded. But at frequencies of 1 gigahertz or greater, they worked just fine — oxide trapping was no longer a hindrance. “When we operate these devices at really high frequency, we noticed that the performance is really good,” she says. “They’re competitive with silicon technology.”
Cai hopes her team’s discovery will give researchers new reason to pursue InGaAs-based computer transistors. The work shows that “the problem to solve is not really the InGaAs transistor itself. It’s this oxide trapping issue,” she says. “We believe this is a problem that can be solved or engineered out of.” She adds that InGaAs has shown promise in both classical and quantum computing applications.
“This [research] area remains very, very exciting,” says del Alamo. “We thrive on pushing transistors to the extreme of performance.” One day, that extreme performance could come courtesy of InGaAs.
This research was supported in part by the Defense Threat Reduction Agency and the National Science Foundation.

Story Source:
Materials provided by Massachusetts Institute of Technology. Original written by Daniel Ackerman. Note: Content may be edited for style and length. More

Lexical borrowing, or the direct transfer of words from one language to another, has interested scholars for millennia, as evidenced already in Plato’s Kratylos dialogue, in which Socrates discusses the challenge imposed by borrowed words on etymological studies. In historical linguistics, lexical borrowings help researchers trace the evolution of modern languages and indicate cultural contact between distinct linguistic groups — whether recent or ancient. However, the techniques for identifying borrowed words have resisted formalization, demanding that researchers rely on a variety of proxy information and the comparison of multiple languages.
“The automated detection of lexical borrowings is still one of the most difficult tasks we face in computational historical linguistics,” says Johann-Mattis List, who led the study.
In the current study, researchers from PUCP and MPI-SHH employed different machine learning techniques to train language models that mimic the way in which linguists identify borrowings when considering only the evidence provided by a single language: if sounds or the ways in which sounds combine to form words are atypical when comparing them with other words in the same language, this often hints to recent borrowings. The models were then applied to a modified version of the World Loanword Database, a catalog of borrowing information for a sample of 40 languages from different language families all over the world, in order to see how accurately words within a given language would be classified as borrowed or not by the different techniques.
In many cases the results were unsatisfying, suggesting that loanword detection is too difficult for machine learning methods most commonly used. However, in specific situations, such as in lists with a high proportion of loanwords or in languages whose loanwords come primarily from a single donor language, the teams’ lexical language models showed some promise.
“After these first experiments with monolingual lexical borrowings, we can proceed to stake out other aspects of the problem, moving into multilingual and cross-linguistic approaches,” says John Miller of PUCP, the study’s co-lead author.
“Our computer-assisted approach, along with the dataset we are releasing, will shed a new light on the importance of computer-assisted methods for language comparison and historical linguistics,” adds Tiago Tresoldi, the study’s other co-lead author from MPI-SHH.
The study joins ongoing efforts to tackle one of the most challenging problems in historical linguistics, showing that loanword detection cannot rely on mono-lingual information alone. In the future, the authors hope to develop better-integrated approaches that take multi-lingual information into account.

Story Source:
Materials provided by Max Planck Institute for the Science of Human History. Note: Content may be edited for style and length. More