Welcome to the November 16, 2022, 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."
|
|
|
Frontier Keeps Top Supercomputer Spot
HPCwire
November 14, 2022
The Frontier system at the U.S. Department of Energy's Oak Ridge National Laboratory remained the world's top supercomputer on the 60th Top500 list, clocking 1.102 exaflops and scoring highest on the HPL-MxP benchmark with 7.9 exaflops. The third-highest-ranking supercomputer, Finland’s Lumi, had the second-highest HPL-MxP score with 2.2 exaflops, while Japan's second-ranked Fugaku supercomputer earned the third-highest HPL-MxP position with 2.0 exaflops. Fugaku held its lead position on the High Performance Conjugate Gradients benchmark with 16 Linpack petaflops, while Frontier scored second with 14 petaflops. Lumi grew its computational footprint, raising its score from 151.9 to 309.1 petaflops.
|
C++ Overtakes PHP, but JavaScript, Python, Java Still Rule
ZDNet
Liam Tung
November 15, 2022
GitHub's 2022 Octoverse report lists JavaScript, Python, Java, TypeScript, and C# as the most-used programming languages in GitHub projects, while PHP was dislodged from sixth place by C++. Hashicorp Configuration Language and Rust were the fastest-growing languages, expanding 56% and more than 50% from last year, respectively. Use of TypeScript grew 37.8%, while use of Lua, Go, Shell, Makefile, C, Kotlin, and Python also saw significant gains. GitHub credits the growth of the Google-maintained Go language to cloud development, as well as projects such as Docker and Kubernetes. More than 94 million developers currently use GitHub to host code, file pull requests, and commits; the platform's community expanded 27% year over year with the addition of 20.5 million users.
|
Cyber Vulnerability in Networks Used by Spacecraft, Aircraft, Energy Generation Systems
University of Michigan News
Zachary Champion
November 15, 2022
Researchers at the University of Michigan and the U.S. National Aeronautics and Space Administration (NASA) discovered a cyberattack that exploits networks used by aircraft, spacecraft, energy generation systems, and industrial control systems. The PCspooF exploit targets the time-triggered ethernet (TTE) system, which lowers costs in high-risk settings by allowing mission-critical and less-critical devices to operate on the same network hardware. PCspoof mimics switches in TTE networks to send out malicious synchronization messages masked by electromagnetic interference. The disruption gradually causes time-sensitive messages to be dropped or delayed, with potentially disastrous effects. The researchers said the exploit can be prevented by replacing copper Ethernet cables with fiber-optic cables, or by deploying optical isolators between switches and untrusted devices.
|
How Big Data Is Transforming Football
Nature
David Adam
November 15, 2022
Data analysis is being incorporated more deeply into football (known in the U.S. as soccer), with organizers at the World Cup tournament in Qatar to pick or bench players via data-informed performance breakdowns. Technologies like motion sensor-equipped vests, global positioning system tracking, and camera surveillance collect a wide array of data at matches, while teams are tapping mathematicians, data scientists, and physicists to extract sense from this information. Researchers at Google's DeepMind and the U.K.-based Liverpool Football Club developed a model to predict player movements in filmed matches, which the club's Ian Graham said can offer insights into tactics.
|
Ensuring AI Works with the Right Dose of Curiosity
MIT News
Rachel Gordon
November 10, 2022
An algorithm designed by Massachusetts Institute of Technology (MIT) researchers surmounts the problem of artificial intelligence (AI) being overly curious and driven to distraction by a task at hand. The algorithm automatically boosts curiosity when necessary, and restrains it if environmental supervision is sufficient for the AI to complete its task. The algorithm successfully performed both difficult and easy exploration tasks when tested on more than 60 videogames, while previous algorithms could only complete either hard or easy tasks. MIT's Zhang-Wei Hong said the algorithm "removes the burden of tuning the balance of exploration and exploitation. Previously what took, for instance, a week to successfully solve the problem, with this new algorithm, we can get satisfactory results in a few hours."
|
Discovery Helps Computers Draw Intricate 2D Animations
Scientific American
Lyndie Chiou
December 1, 2022
Chenxi Liu and colleagues at Canada's University of British Columbia devised the ConTesse algorithm to solve the hidden line problem of computer graphics and enable computers to compose complex two-dimensional (2D) animations of three-dimensional (3D) models. ConTesse addresses the challenge by correcting the model's outline or contour, rather than a mesh of triangular tiles. The algorithm traces a 3D configuration's edges with line segments, then squashes this approximate contour into 2D and attempts to tile its interior with a mesh. It modifies that part of the contour wherever the inside mesh mistakenly crosses over itself, then regenerates the mesh using the fixed contour and projects it back onto the 3D object for a final visibility review.
|
Italy Outlaws Facial Recognition Tech, Except to Fight Crime
Reuters
Elvira Pollina; Federico Maccioni
November 14, 2022
Italy has banned the use of facial recognition technology and "smart glasses," except when such technologies play a role in judicial investigations or the fight against crime. The country's Data Protection Agency issued a rebuke to two municipalities experimenting with the technologies, and said facial recognition systems using biometric data will not be allowed until a specific law is adopted, or at least until the end of next year. Under EU and Italian law, the processing of personal data by public bodies using video devices is generally allowed on public interest grounds, and when linked to the activity of public authorities. However, the agency noted, municipalities that want to use such devices will have to secure "urban security" agreements with central government representatives.
|
Hard-to-Crack Hardware
King Abdullah University of Science and Technology (Saudi Arabia)
November 14, 2022
A team of researchers at Saudi Arabia's King Abdullah University of Science and Technology (KAUST) has created an integrated circuit logic lock that could advance cyberattack-resistant electronic devices. The researchers based the logic lock on a magnetic tunnel junction (MTJ), which uses spintronics to function. The MTJ's electronic output relies on the spin alignment of the electrons inside it, and only generates the correct output for the circuit when it receives the appropriate key signal input. KAUST's Yehia Massoud said, "With the advancement in fabrication methods, the possibility of using emerging spintronic device structures in the chip design has increased. These properties make spintronic devices a potential choice for exploring hardware security."
|
Amazon Unveils Smaller Delivery Drone That Can Fly in Rain
Bloomberg
Spencer Soper; Matt Day
November 10, 2022
Retailing giant Amazon.com has introduced a new delivery drone that can fly through light precipitation. Scheduled to go into service in 2024, the MK30 drone will replace the MK27-2, which will be used to deliver purchases in Lockeford, CA, and College Station, TX, this year. Amazon listed MK30 upgrades that include a longer flight range, tolerance for a broader range of temperatures, and new safety features. The drone is also smaller and produces less noise than the MK27-2. The MK30 is Amazon's latest attempt to realize owner Jeff Bezos' vision of autonomous drones that can transport packages weighing less than five pounds to customers as soon as a half-hour after an order is placed.
*May Require Paid Registration
|
Reducing Redundancy to Accelerate Complicated Computations
Thomas Jefferson National Accelerator Facility
November 15, 2022
Scientists at the U.S. Department of Energy's Thomas Jefferson National Accelerator Facility and the College of William & Mary have developed a tool to optimize supercomputing time. Their MemHC framework structures the memory of a graphics processing unit (GPU) to accelerate the calculation of many-body correlation functions. The researchers created three memory management methods that reduce redundant memory operations and expedite calculation of tensor contractions 10-fold. They coded MemHC to enable memories to persist on the GPU in a manner more appropriate for calculations, reducing the GPU's input and output tasks to concentrate on communication between the GPU and its host central processing unit.
|
How an 'Electronic Nose' Could Help Fight Wildfires
CNN
Jacopo Prisco
November 14, 2022
German startup Dryad is testing an "ultra-early" wildfire warning system across a dozen forest regions worldwide, using a solar-powered sensor equipped with a gas detector. Dryad's Carsten Brinkschulte said this "electronic nose" can detect fires by sensing hydrogen, carbon monoxide, and volatile organic compounds that make up smoke in the air. Once the sensor detects a fire, its antenna transmits a wireless signal to other devices, which is routed to the Internet via satellite and 4G connections before being sent to forest managers. Brinkschulte said Dryad also issues alerts and can link directly with the local fire brigade's information technology systems, supplying "the exact GPS [global positioning system] coordinates of the sensor that picked up the fire."
|
AI Uses Artificial Sleep to Learn New Task without Forgetting the Last
New Scientist
Jeremy Hsu
November 10, 2022
Scientists at the University of California, San Diego (UCSD) and the Czech Republic's Czech Academy of Sciences taught an artificial intelligence (AI) to learn a second distinct task without overwriting connections learned from a first task, through the use of simulated sleep. UCSD's Erik Delanois said it was critical to "have rapidly alternating sessions of training and sleep" while the AI was learning the second task, which consolidated links from the first task that would have otherwise been forgotten. "Such a network will have the ability to combine consecutively learned knowledge in smart ways, and apply this learning to novel situations—just like animals and humans do," said the University of Massachusetts Amherst's Hava Siegelmann.
|
Microscopy, Simulation Unite to Improve New-Age Polymers
Stanford Engineering
Andrew Myers
November 11, 2022
Stanford University researchers used electron microscopy to visualize the real-world arrangement of molecules in a new-age polymer material, and computer modeling to simulate how certain structural changes in the material could improve the flow of electricity. Said Stanford's Alberto Salleo, “By revealing the key connection between a polymer’s structure and its electrical function, this ability to visualize real microstructures and tinker with them mathematically offers new and powerful insight into the design of promising new materials." The insights the team learned through its approach improve understanding of how the arrangement, coupling, and geometries of chains of molecules across the mesoscale improve, or inhibit, electrical transport through bulk polymers. “These simulated chains can be modified on the computer in ways that are difficult, if not impossible, in the real world,” said study co-author Andrew Spakowitz.
|
|
