Welcome to the April 19, 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."
|
|
Papworth Honored with ACM Breakthrough in Computing Award
ACM April 19, 2023
David B. Papworth has been named to receive the ACM Charles P. "Chuck" Thacker Breakthrough in Computing Award for his work on microprocessors at Intel. Papworth was a lead designer of the Intel P6, whose microarchitectural paradigm involved decomposing complex x86 instructions into sequences of micro-operations that flowed through a micro data flow engine. ACM President Yannis Ioannidis said, "The introduction of Intel's P6 microarchitecture in 1995 was an important milestone during a time when the personal computing software and hardware industry really started to take off. As the lead developer of the P6 microarchitecture, David Papworth was one of the unsung heroes of the decade and his contributions are still in use today."
|
Oklahoma Businesses Recruit Tech Workers from Ukraine
The Wall Street Journal Alicia A. Caldwell April 16, 2023
A nonprofit based in Tulsa, OK, is recruiting Ukrainians fleeing the Russian invasion for jobs in the region's expanding technology sector. Local officials and business leaders say the recent arrivals are appealing because they qualify to circumvent immigration restrictions under the Biden administration's Uniting for Ukraine program, which allows American sponsors to sponsor Ukrainians on two-year work visas. The inTulsa Visa Network runs the Ukrainian program as an offshoot of the inTulsa tech-sector recruitment organization, which brings in workers from other U.S. regions. The group has cleared 15 Ukrainians and their families to relocate to Tulsa, of whom six have found local employment.
*May Require Paid Registration
|
Optimization Could Cut AI Training’s Carbon Footprint by 75%
University of Michigan Computer Science and Engineering April 17, 2023
The Zeus open source optimization framework developed by the University of Michigan's Mosharaf Chowdhury and colleagues examines deep learning models during training to determine the best balance between energy use and training speed. The researchers say Zeus could slash training's energy consumption by up to 75% without changing system hardware by tuning the graphics processing unit (GPU) power limit and the deep learning model's batch size parameter in real time. The researchers were able to visualize the optimal energy-training time tradeoff point by showing every possible blending of these parameters. They also developed a software package they named Chase, which prioritizes speed when low-carbon energy is available and opts for efficiency while reducing speed during peak times.
|
Entangled Photons Produced Entirely On-Chip
IEEE Spectrum Edd Gent April 17, 2023
A photonic chip developed by researchers at Germany's Leibniz University of Hannover integrates the key components needed to produce entangled photon pairs. The hybrid chip incorporates two different semiconductor technologies to generate laser light and convert it into high-quality entangled photons. One section of the chip features an indium phosphide optical amplifier; it connects to the other section, containing three silicon nitride "micro-ring resonators." The first two rings filter noise from the laser signal, and the third ring generates a pair of entangled photons. The researchers said the chip can generate 8,200 pairs of entangled photons per second at optical communications wavelengths, while consuming only 3 watts of energy.
|
A Trick of the Hat
University of Waterloo News (Canada) Joe Petrik April 13, 2023
A group of British, Canadian, and American researchers has proven the existence of a 13-sided shape called an aperiodic monotile or einstein, also known as "the hat," which can fill an infinite plane free of overlaps or gaps in a pattern that never repeats and can never be made to repeat. Retired print technician David Smith came across the shape last fall and explored it with paper cut-outs, before asking Craig S. Kaplan at Canada's University of Waterloo to run the hat through software he had developed to better characterize its shape. Kaplan said the software helped elucidate things about the hat's generic tilings of which the researchers had no knowledge. He said they found “The hat is not just a single shape that tiles aperiodically, but a member of a continuum of shapes,” adding, “the more interesting question is, are there fundamentally aperiodic monotiles?”
|
Waiting for Quantum Computers, Software Engineers Get Creative
Reuters Jane Lee April 17, 2023
Startups are developing a new type of software inspired by quantum algorithms that can be used on classical computers with graphic processing units (GPUs) as development of quantum computers continues. QC Ware's Matt Johnson said the firm used Nvidia GPUs to offer clients "a big step-change in performance" while providing "a bridge to quantum processing in the future." QC Ware's quantum-inspired software platform Promethium uses GPUs to simulate chemical molecules on classical computers. QC Ware's Robert Parrish said the platform can reduce simulation times to minutes for molecules up to 100 atoms, and to hours for molecules up to 2,000 atoms. Meanwhile, Alphabet spinoff SandBoxAQ offers a quantum-inspired algorithm that facilitates biopharma simulations using Google's Tensor Processing Unit (TPU) artificial intelligence chip.
|
Autonomous Bus Sounds: All About When, Not How
Cornell Chronicle Patricia Waldron April 17, 2023
Cornell University's Malte Jung and Hannah Pelikan from Sweden's Linköping University are using sound to improve autonomous buses' navigation and communication capabilities. The researchers found the timing of sounds is the critical factor in ensuring effective social engagement by autonomous buses in the Swedish town of Linköping. They played various sounds through a Bluetooth speaker outside of a bus to alert pedestrians and cyclists to the vehicle’s approach, analyzed videos of interactions, and chose new sounds to test based on that information. Jung and Pelikan observed through video analysis that timing and duration, rather than sound type, were key to signaling the bus's intentions.
|
Quadruped Robot Uses Satellite Tools to Walk a Balance Beam
New Atlas Ben Coxworth April 17, 2023
Carnegie Mellon University engineers have developed a device that could improve the balance of quadruped robots. The engineers mounted two reaction wheel actuators (RWAs), generally used to adjust the angular momentum of satellites, to the back of a Unitree A1 robot. Housed in a single module, one RWA controls the robot's pitch axis while the other controls its roll access, automatically compensating for changes in the robot's center of balance regardless of which feet are touching the beam. The device made it possible for the robot to walk along a 9.8-foot by 2.4-inch balance beam without tipping over and allowed it to flip around in mid-air and land on its feet when dropped from a height of 1.6 feet.
|
3D Printing Helps Study of Plants
PBS North Carolina Frank Graff April 18, 2023
Scientists at North Carolina State University (NC State) three-dimensionally (3D)-bioprinted cells from Arabidopsis Thaliana and soybean plants. NC State's Lisa Van den Broeck said the process uses living plant cells as "bioink," with an ultraviolet filter maintaining environmental sterility and multiple print heads extruding bioinks concurrently. The researchers bioprinted plant cells without cell walls, along with nutrients, growth hormones, and a thickening agent made from seaweed. NC State's Ross Sozzani said the study found "that 3D bioprinting can be useful to study cellular regeneration in crop plants. It also shows the potential of using 3D bioprinting to identify the optimal compounds needed to support plant cell viability in a controlled environment."
|
Location Intelligence Shines Light on Disinformation
Oak Ridge National Laboratory April 14, 2023
Researchers in the Oak Ridge National Laboratory's (ORNL) Location Intelligence Group are studying the flow of disinformation online using an automated method for quantifying social media users' intent. In a test of 4.7 million COVID-19-related tweets from over 14,000 users, the researchers determined the approach can correlate breaking news notifications with online responses from disinformation actors. The approach factors in the unique spatial patterns of information spread and how social media users increase this spread.
|
Apple's Macs May No Longer Escape Ransomware
Ars Technica Lily Hay Newman April 18, 2023
Security researchers are analyzing newly discovered Mac ransomware samples from the Russia-based LockBit gang, the first known instance of a major ransomware group tinkering with macOS versions of its malware. The samples seem to have first appeared in the VirusTotal malware analysis repository in November/December 2022, but were only noticed on April 17. One sample appears to be a version of the encryptor targeting newer Macs running Apple processors and older Macs driven by PowerPC chips. The Objective-See Foundation's Patrick Wardle said, “In some sense, Apple is ahead of the threat, as recent versions of macOS ship with a myriad of built-in security mechanisms aimed to directly thwart, or at least reduce the impact of, ransomware attacks. However, well-funded ransomware groups will continue to evolve their malicious creations.”
|
AI Helps Cyclists Work Out How Much to Eat During Tour de France
New Scientist Matthew Sparkes April 18, 2023
Elite cyclists are using artificial intelligence to plan their caloric intake during races like the Tour de France. Researchers at Maastricht University in the Netherlands and Dutch professional cycling organization Team Jumbo-Visma used machine learning and mathematical techniques to better formulate riders' diets. They compiled a statistical model using data from previous races, including each rider's body measurements and power output, the route and elevation of the race stages, weather, and wind direction. The researchers applied this model to calculate calorie requirements for any rider on any stage route. The model was found to be more accurate than coaches in predicting these requirements for previous stages in the Tour de France and Italy's Giro d'Italia from 2019.
|
|