Welcome to the August 11, 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."
|
|
|
White House Launches AI-Based Contest to Secure Government Systems from Hacks
Reuters
Zeba Siddiqui
August 9, 2023
The White House announced the launch of a competition to encourage the use of artificial intelligence (AI) to pinpoint and correct vulnerabilities in U.S. government infrastructure. The Defense Advanced Research Projects Agency (DARPA) will administer the two-year contest, which offers about $20 million in prizes; leading AI technology vendors Google, Anthropic, Microsoft, and OpenAI will provide systems for the competition. Deputy national security adviser for cyber and emerging technology Anne Neuberger said the goal of the competition “is to catalyze a larger community of cyber defenders who use the participating AI models to race faster—using generative AI to bolster our cyber defenses."
|
Indiana Tests if the Heartland Can Transform into a Chip Hub
The New York Times
Cecilia Kang; Ana Swanson
August 7, 2023
State leaders aim to transform Indiana into a center of microchip manufacturing and research as a test case of the Biden Administration's $52-billion CHIPS (Creating Helpful Incentives to Produce Semiconductors) and Science Act. Indiana-based companies and universities have applied for multiple CHIPS Act grants, and the state's advantages include abundant land for chip factories and water to cool their equipment and rinse silicon wafers. Purdue University's engineering school also has pledged to train technicians and scientists for chip production, and earlier this year Minneapolis-based chipmaker SkyWater committed to investing $1.8 billion to build a chip factory next to the university. Critical to the state's ambitions is the LEAP Innovation District in the town of Lebanon, which has offered incentives to chipmakers like Taiwan's TSMC and South Korea's SK Hynix.
*May Require Paid Registration
|
What Do a Jellyfish, a Cat, a Snake, and an Astronaut Have in Common? Math
Caltech News
August 9, 2023
A team led by California Institute of Technology (Caltech) researchers developed an algorithm that describes the various motions that help animals navigate their environments by changing their body shape. The principle of least dissipation can be used to explain these motions: natural systems always seek to be as efficient as possible. This principle was used as the starting point in developing a computer model of the different kinds of motions. Said Caltech's Peter Schröder, "It's just beautiful that you can identify a fairly simple governing principle of a whole class of different kinds of motion. It's not 100% accurate but shows remarkable agreement with motion observed in real life, suggesting that it captures a major part of what happens in nature.”
|
Using Social Media to Raise Awareness of Women's Resources
MIT News
Peter Dizikes
August 7, 2023
Researchers at the Massachusetts Institute of Technology, the Egyptian Center for Women’s Rights, and Mexico's Instituto Tecnologico Autonomo de Mexico conducted a randomized experiment in Egypt to gauge awareness of women's issues through social media. The researchers sent social media-posted videos to women recruited via Facebook as reminders to watch TV programming from an Egyptian human rights lawyer discussing gender norms and violence; as a result, the women consumed more media content about the issue, learned more about the resources available to them, and reportedly used some resources more in response to violence among women. However, the researchers found the women’s long-term views about gender/marital equity and sexual violence remained largely unchanged.
|
Downfall Attacks on Intel CPUs Steal Encryption Keys, Data
BleepingComputer
Ionut Ilascu
August 8, 2023
Google's Daniel Moghimi exploited the so-called "Downfall" bug in Intel central processing units to steal passwords, encryption keys, and private data from computers shared by multiple users. The transient execution side-channel vulnerability affects multiple Intel microprocessor lines, allowing hackers to exfiltrate Software Guard eXtensions-encrypted information. Moghimi said Downfall attacks leverage the gather instruction that "leaks the content of the internal vector register file during speculative execution." He developed the Gather Data Sampling exploit to extract AES 128-bit and 256-bit cryptographic keys on a separate virtual machine from the controlled one, combining them to decrypt the information in less than 10 seconds. Moghimi disclosed the flaw to Intel and worked with the company on a microcode update to address it.
|
Researchers Improve Environmental Monitoring Applications
Wayne State University
August 9, 2023
Researchers at Wayne State University (WSU) and the University of Delaware have enhanced vapor intrusion monitoring by integrating the Internet of Things sensor network with Edge Computing (IoTEC). WSU's Yongli Wager explained, "We found that IoTEC-based monitoring revealed a reduction in unnecessary data transmission and data latency, or how long it takes for data to be stored or retrieved. In addition, using IoTEC resulted in an estimated cost reduction of 55-82% for vapor intrusion monitoring that covered five houses." The researchers also monitored and forecast system performance of wastewater-based algae cultivation using machine learning tools with IoTEC, improving data processing and analysis by deploying different algorithms at edge servers.
|
NOAA Completes Upgrade to Weather, Climate Supercomputer System
National Oceanic and Atmospheric Administration
August 10, 2023
The U.S. National Oceanic and Atmospheric Administration (NOAA) and the Department of Commerce have completed upgrading the Weather and Climate Operational Supercomputing System with a 20% capacity expansion. NOAA's total supercomputing capacity is now 49 petaflops with the addition of twin supercomputers in Manassas, VA, and Phoenix, AZ, that run at 14.5 petaflops each. NOAA's Ken Graham said, "This increased supercomputing power allows for upgrades to specific modeling systems that will help weather forecasters deliver more accurate weather forecasts, watches, and warnings and improved certainty in a forecast." The upgrade will improve the U.S. Global Forecast System's horizontal resolution from its current 13 kilometers (eight miles) to nine kilometers (5.5 miles). The Global Ensemble Forecast System also will better simulate emissions like wildfire smoke, dust, and fog by capturing radiatively active aerosols more accurately.
|
Nvidia Unveils Faster Chip to Cement AI Dominance
Bloomberg
Ian King
August 8, 2023
Nvidia unveiled the Grace Hopper Superchip (GH200) at the ACM SIGGRAPH conference to solidify the technology company's lead in the artificial intelligence (AI) accelerator market with updated speed and capacity. Nvidia said the combined graphics chip/processor's high-bandwidth memory 3 (HBM3e) can access data at 5 terabytes per second, and will go into production in the second quarter of next year. CEO Jensen Huang envisions accelerator chips supplanting conventional datacenter equipment, with the GH200 forming the core of a new server architecture that can accommodate and more quickly access greater volumes of information. Nvidia said deploying two Superchips together in servers offers more than 3.5 times the capacity of a current model.
*May Require Paid Registration
|
Sensor Technology Discovery to Combat Water Contamination, More
Argonne National Laboratory
Joseph E. Harmon
August 10, 2023
Researchers at the U.S. Department of Energy's Argonne National Laboratory and the universities of Chicago and Wisconsin-Milwaukee have created a methodology for mass-producing sensors that can detect contaminants in tap water. The sensors use a nanometer-thick graphene coating on a silicon substrate with imprinted gold electrodes and a nanometer-thick insulating layer of aluminum oxide. The researchers customized each sensor to detect either lead, mercury, or E. coli. Using machine learning algorithms to analyze the results, they were able to quantify toxin levels down to the parts per billion, even in the presence of interfering elements. Said Argonne’s Junhong Chen, "You can combine three, 30, or 300 sensors, with each tailored to detect different constituents."
|
VR Headsets Are Vulnerable to Hackers
UC Riverside News
David Danelski
August 8, 2023
Computer scientists at the University of California, Riverside found hackers can translate the movements of virtual reality (VR) and augmented reality (AR) headset users into words using spyware and artificial intelligence. In one example, spyware used a headset user's motions to record their Facebook password as they air-typed it on a virtual keyboard. Spies also could potentially access a user's actions during virtual meetings involving confidential information by interpreting body movements. One exploit showed hackers retrieving a target's hand gestures, voice commands, and keystrokes on a virtual keyboard with over 90% accuracy. Researchers also developed a system called TyPose that uses machine learning to extract AR/VR users' head motions to deduce words or characters they are typing.
|
Incorporating Human Error into Machine Learning
University of Cambridge (U.K.)
August 10, 2023
Scientists at the U.K.'s University of Cambridge, The Alan Turing Institute, Princeton University, and Google DeepMind are incorporating uncertainty into machine learning (ML) systems. The researchers utilized established image classification datasets so humans could supply feedback and rate their uncertainty level when annotating specific images. They learned the systems can handle uncertain feedback better when training with uncertain labels, although their overall performance degrades rapidly with human feedback. Cambridge's Matthew Barker said, "We're trying to bridge [behavioral research and ML] so that machine learning can start to deal with human uncertainty where humans are part of the system."
|
Algorithm Creates Shapes That Roll Down Pre-Determined Paths
The Register (U.K.)
Lindsay Clark
August 9, 2023
An algorithm created by researchers at South Korea's Institute for Basic Science and Switzerland's University of Geneva can be used to assemble objects that roll down pre-determined paths with potential quantum-mechanical and medical applications. The researchers had to design shapes that followed a pre-determined path, ending at the same point on the object's surface where it began. They achieved this by engineering "two-period trajectoids" that spin twice over a path to handle nearly every imaginable trajectory. Scientists at the Georgia Institute of Technology and Oklahoma State University suggested potential uses for trajectoids, including controlling the evolution of quantum bits and mitigating noise in medical magnetic resonance imaging by controlling spin dynamics.
|
SPAD Detectors Achieve 3D Quantum Ghost Imaging
Optica
August 8, 2023
Scientists in Italy and Germany have acquired the first three-dimensional measurements with a computational imaging approach called quantum ghost imaging. The researchers said their method utilizes single-photon avalanche diode (SPAD) array detectors, which possess multiple independent pixels with dedicated timing circuitry recording each pixel's picosecond-resolution detection time. The team was able to reconstruct the photonic entanglement and object depth by comparing the detection time of every pixel with the detection of the single-pixel detector. Carsten Pitsch at Germany's Fraunhofer Institute of Optronics, System Technologies, and Image Exploitation said the technique "can be applied to image materials and tissues that are sensitive to light or drugs that become toxic when exposed to light, without any risk of damage."
|
|
