Welcome to the November 17, 2023, edition of ACM TechNews, providing timely information for IT professionals three times a week.
ACM TechNews mobile apps are available for Android phones and tablets (click here) and for iPhones (click here) and iPads (click here).
To view "Headlines At A Glance," hit the link labeled "Click here to view this online" found at the top of the page in the html version.
The online version now has a button at the top labeled "Show Headlines."
|
|
|
Gordon Bell Prize Awarded for Materials Simulations That Achieve Quantum Accuracy at Scale
Association for Computing Machinery
November 16, 2023
A team of researchers from U.S. and Indian institutions was awarded the 2023 ACM Gordon Bell Prize for a materials-simulation framework that combines the accuracy provided by Quantum Many-Body (QMB) methods with the efficiency of Density-Functional Theory (DFT). The team proposed three interconnected modules for their new method. One was a methodological advance in inverse DFT (invDFT) linking QMB methods to DFT. The second module was a machine-learned density functional trained with invDFT data, commensurate with quantum accuracy, which they dubbed MLXC. The third was an adaptive higher-order spectral finite-element based DFT implementation that integrates MLXC with efficient solver strategies and supercomputing innovations in linear algebra, mixed-precision algorithms, and asynchronous compute-communication. The award was presented at the International Conference for High-Performance Computing, Networking, Storage and Analysis (SC23).
|
Printing Robots with Bones, Ligaments, Tendons
ETH Zurich (Switzerland)
Fabio Bergamin
November 15, 2023
Researchers at Eidgenössische Technische Hochschule (ETH) Zürich and U.S. startup Inkbit, a spinoff of the Massachusetts Institute of Technology (MIT), three-dimensionally (3D) printed a robotic hand with bones, ligaments, and tendons using a new laser scanning technique. The advance was made possible due to the advent of slow-curing polymers in the 3D printing process. The researchers used laser scanning to immediately check each printed layer for any surface irregularities. Explained MIT's Wojciech Matusiky, “A feedback mechanism compensates for these irregularities when printing the next layer by calculating any necessary adjustments to the amount of material to be printed in real time and with pinpoint accuracy."
|
DeepMind Accurately Forecasts Weather on a Desktop Computer
Nature
Carissa Wong
November 14, 2023
Google DeepMind developed a machine-learning weather-forecasting model that outperformed the best conventional forecasting tools, as well as other artificial intelligence (AI)-based approaches. The GraphCast model can run on a desktop computer and makes its predictions in minutes. Researchers trained the model using estimates of past global weather made from 1979 to 2017 by physical models, allowing GraphCast to learn links between different weather variables. The trained model uses the current state of global weather, and weather estimates from six hours earlier, to predict the weather six hours ahead. The researchers found GraphCast could use global weather estimates from 2018 to make forecasts up to 10 days ahead in less than a minute, with the resulting predictions more accurate than those made by the European Centre for Medium-Range Weather Forecasts’ High RESolution forecasting system, which takes hours to forecast.
|
Brain-like Computing System More Accurate with Custom Algorithm
University of California, Los Angeles Newsroom
Wayne Lewis
November 13, 2023
A training algorithm developed by researchers at the University of Sydney (USyd) in Australia helped an experimental computing system physically modeled after the biological brain “learn” to identify handwritten numbers with an overall accuracy of 93.4%. Researchers at the University of California, Los Angeles have been working on the new platform technology for computation for 15 years. The technology is a brain-inspired system composed of a tangled-up network of nanowires containing silver, laid on a bed of electrodes. The system receives input and produces output via pulses of electricity. Collaborators at USyd developed a streamlined algorithm for providing input and interpreting output. The algorithm is customized to exploit the system’s brain-like ability to change dynamically and to process multiple streams of data simultaneously.
|
High-Tech Soccer Ball Unveiled for Euro 2024
Associated Press
November 15, 2023
A high-tech soccer ball that helps with offside decisions will make its European Championship debut next year after being used at last year's World Cup. European soccer governing body UEFA and manufacturer Adidas unveiled the ball for Euro 2024 in Berlin on Wednesday. A chip fixed on a gyroscope inside the ball sends data 500 times per second to record the point at which it is kicked. The “kick point” helps match officials make offside decisions using multiple camera angles to create three-dimensional visualizations that illustrate player movement. UEFA said the “connected ball technology gives unprecedented insight into every element of the movement of the ball and contributing to UEFA’s video assistant refereeing decision-making process."
|
Twist on AI Makes the Most of Sparse Sensor Data
Los Alamos National Laboratory News
November 13, 2023
Los Alamos National Laboratory researchers have developed an approach to artificial intelligence (AI) that enables the reconstruction of a broad field of data from a small number of sensors using low-powered “edge” computing. The AI technique, dubbed Senseiver, builds on a model called Perceiver IO developed by Google by applying the techniques of natural language models to the problem of reconstructing information about a broad area from relatively few measurements. The team applied the model to a National Oceanic and Atmospheric Administration sea-surface-temperature dataset. By integrating measurements taken over decades from satellites and sensors on ships, the model was able to forecast temperatures across the entire body of the ocean. Said Los Alamos' Dan O’Malley, "Using fewer parameters and less memory requires fewer central processing unit cycles on the computer, so it runs faster on smaller computers."
|
Hologram Lets Philippines’ Marcos Speak in Singapore While Visiting U.S.
Bloomberg
Olivia Poh; Andreo Calonzo
November 15, 2023
About an hour after delivering a speech in California on Wednesday, Philippine President Ferdinand Marcos Jr. appeared in Singapore via hologram. As he appeared in the middle of a purplish vortex animation, Marcos said the Southeast Asian nation will prioritize digitalizing cash transactions. “We will keep our fintech ecosystem open to collaboration with a wide array of stakeholders, both domestically and internationally,” said Marcos, who was attending the Asia-Pacific Economic Cooperation meeting in San Francisco.
|
NASA Supercomputers Reveal Surface Effects of Rocket Exhaust on the Moon
National Aeronautics and Space Administration
Gianine Figliozzi
November 14, 2023
National Aeronautics and Space Administration (NASA) researchers developed software tools to predict the surface effects of the supersonic plumes of hot gas emitted when a rocket lands or takes off. Using NASA's Pleaides supercomputer, the team produced a simulation of the Apollo 12 lander’s engine plumes interacting with the Moon's surface and the predicted erosion closely matched what happened during landing. The simulation depicts the last half-minute of descent before engine cut-off, showing the predicted forces exerted by plumes on a flat computational surface. The software tools are already being used to predict cratering and visual obscuration on upcoming missions to the Moon.
|
Sensory System Identifies Textures
Interesting Engineering
Jijo Malayil
November 14, 2023
Researchers led by Chuan Fei Guo at China’s Southern University of Science and Technology developed an artificial sensory system that can detect tiny textures such as twill, corduroy, and wool, with high resolution. The team devised a flexible slip sensor replicating human fingerprint characteristics, allowing the system to detect microscopic details on surface textures when touching or sliding the sensor across a surface. They fitted the sensor onto a prosthetic human hand using machine learning. The team detailed how a real-time sensory system can be used to classify 20 different textiles with an average identification accuracy of 98.6%, as well as how the findings can be shown in a visual interface.
|
Ukraine’s ‘Secret Weapon’ Against Russia Is a U.S. Tech Company
Time
Vera Bergengruen
November 14, 2023
U.S. facial recognition company Clearview AI has become Ukraine’s “secret weapon" in its war against Russia. More than 1,500 officials across 18 Ukrainian government agencies are using its technology, which has helped them identify more than 230,000 Russian soldiers and officials who have participated in the Russian invasion. Ukraine also relies on the company to assist with other tasks, including processing citizens who lost their identification and locating abducted Ukrainian children. Ukraine has run at least 350,000 searches of Clearview’s database in the 20 months since the outbreak of the war. Said Clearview AI CEO Hoan Ton-That, “Using facial recognition in war zones is something that's going to save lives.”
|
Autonomous Lab Discovers Best-in-Class Quantum Dot in Hours
North Carolina State University News
Matt Shipman
November 13, 2023
A team led by researchers at North Carolina State University (NC State) developed an autonomous system to address a challenge regarding enhancing properties of materials called perovskite quantum dots via “doping.” Doped quantum dots are semiconductor nanocrystals that have been infused with specific impurities to alter their optical and physicochemical properties in a bid to create photonic and optoelectronic devices. The SmartDope system is a “self-driving” lab that starts with the researchers telling it which precursor chemicals to work with and providing it a designated goal. “As SmartDope collects data on each of its experiments, it uses machine learning to update its understanding of the doped quantum dot synthesis chemistry and inform which experiment to run next, with the goal of making the best quantum dot possible,” explained NC State's Milad Abolhasani.
|
Cloud Data Storage Security Approach Taps Quantum Physics
AIP Publishing
November 14, 2023
Yong Zhao of Chinese quantum information technology company QuantumCTek and colleagues developed a secure cloud data storage method that uses quantum random numbers as encryption keys and disperses them via Shamir’s secret sharing algorithm. Shamir’s algorithm involves distributing private information to a group so “the secret” will be revealed only when a majority pools their knowledge. It’s common to combine quantum key distribution (QKD) and Shamir’s secret sharing algorithm for secure storage, but solutions can be costly, as they involve significant cloud storage space requirements. The method developed by Zhao's team uses quantum random numbers as encryption keys, dispersing the keys via Sharmir’s secret sharing algorithm, applying erasure coding within ciphertext, and securely transmitting the data through QKD-protected networks to distributed clouds. Said Zhao, “In essence, our solution is quantum-secure and serves as a practical application of the fusion between quantum and cryptography technologies."
|
Dark Matter Data Salvaged from Balloon-borne Telescope That Landed Hard
Space.com
Sharmila Kuthunur
November 14, 2023
Ellen Sirks of the University of Sydney led a multinational team that was able to recover dark matter data from a telescope that launched on a balloon earlier this year, lost communication with Earth, and suffered damage upon landing. The data it had collected included 200 gigabytes' worth of images of galaxy clusters. Usually, data gathered by balloon-borne telescopes are downloaded using a satellite, but rapid downloads require line-of-sight communications. "In our case, we were getting so much data per night that it would just be incredibly slow and expensive to retrieve this data mid-flight," Sirks said. "At the moment, the most efficient way for us to download data is to copy it onto an SD drive and just drop it to Earth."
|
|
