Welcome to the October 21, 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."

People walking toward the Georgia-Russia border crossing at Verkhny Lars. Hundreds of Russia's Top Software Developers May Have Left the Country
New Scientist
Jeremy Hsu
October 19, 2022

Hundreds of top software developers appear to have exited Russia during its invasion of Ukraine, with many leaving before Russian president Vladimir Putin announced a partial military mobilization last month. Nearly 23% of Russian developers who contributed the most to coding projects on the GitHub software development platform changed their location data or deleted their profiles between February 2021 and June 2022. Johannes Wachs at Austria's Vienna University analyzed data of active GitHub developers a year before the invasion and several months into the war. He said both the most active developers and those most central to GitHub are likely to leave. Samuel Bendett at U.S. think tank the Center for New American Security suggested the exodus could seriously undermine Russia's economy or workforce, which had few information technology workers before the exodus.

Full Article

Marcin Copek and Masado Alexander Ishii are recipients of the 2022 ACM-IEEE Computer Science George Michael Memorial HPC Fellowships. Copik, Ishii to Receive 2022 ACM-IEEE CS George Michael Memorial HPC Fellowships
October 19, 2022

ACM and IEEE have named Marcin Copik at Switzerland's ETH Zurich and the University of Utah's Masado Alexander recipients of the 2022 ACM-IEEE Computer Science CS George Michael Memorial HPC Fellowships. Copik incorporated the Function-as-a-Service programming model into high-performance computing (HPC) applications by creating high-performance software and hardware for the serverless stack. Ishii primarily developed the University of Utah's Dendro-KT framework, a series of lightweight, dimension-parameterized, parallel meshing algorithm that also facilitate the development of high-orders and variable order-in-time formulation. Ishii also has helped devise methods and codes for large-scale fluid models around complex objects.

Full Article
How Can Digital Data Stored as DNA Be Manipulated?
CNRS News (France)
October 19, 2022

Researchers at France's CNRS (the French National Center for Scientific Research) and ESPCI Paris-PSL, and Japan's University of Tokyo, leveraged enzymes, artificial neurons, and neural networks to retrieve or manipulate digital data stored as DNA. Using the reactions of three enzymes, the researchers developed chemical neurons that reproduce actual neurons' network architecture and ability to perform complex calculations. These chemical neurons can perform calculations with data on DNA strands and express the results as fluorescent signals. The researchers also enhanced the precision of their calculations using two layers of artificial neurons and the microfluidic miniaturization of reactions, which facilitates tens of thousands of reactions.

Full Article

AI-Powered Material Can Learn Behaviors, Adapt AI-Powered Material Can Learn Behaviors, Adapt
Interesting Engineering
Brittney Grimes
October 19, 2022

Mechanical engineers at the University of California, Los Angeles have developed an artificial intelligence-powered material that learns behaviors over time, and can adjust to changing circumstances. The so-called mechanical neural network (MNN) features a structural system of independently tunable beams arranged in a triangular lattice pattern. The researchers said each beam consists of a "voice coil, strain gauges, and flexures that enable the beam to change its length, adapt to its changing environment in real time, and interact with other beams in the system." An optimization algorithm uses strain-gauge data to calculate rigidity values to govern the network's adaptation, determining how much force should be applied. Cameras on the MNN's outer nodes check the strain-gauge system's validity.

Full Article
Deep Learning with Light
MIT News
Adam Zewe
October 20, 2022

Massachusetts Institute of Technology researchers have reduced computing latency on low-powered connected devices via the Netcast architecture, which shuttles the operation of deep learning models to a central server where components are imprinted onto light waves. Netcast stores weights in the central server, which is linked to a smart transceiver that applies silicon photonics to fetch trillions of weights from memory each second; these weights are encoded onto light waves. Fiber optics route the waves to a connected device, while the receiver uses a Mach-Zehnder modulator to conduct computations using the encoded elements. The modulator performs trillions of multiplications per second, demonstrating a 100-fold improvement in energy efficiency compared to other techniques.

Full Article

A crystal structure surrounds molecular qubit. Stability in Asymmetry: Scientists Extend Qubit Lifetimes
Argonne National Laboratory
Leah Hesla
October 13, 2022

Researchers from the U.S. Department of Energy's Argonne National Laboratory, the Massachusetts Institute of Technology, Northwestern University, the University of Chicago, and the U.K.'s University of Glasgow have extended the lifetime of molecular qubits by disrupting the symmetry of the surrounding crystal's structure. The qubits, comprised of a chromium-based ion attached to carbon-based molecules, are protected from potentially disruptive magnetic fields by the crystal's asymmetry, which enabled the molecular qubits to achieve a coherence time of 10 microseconds, compared to 2 microseconds for those in symmetrical structures. Said University of Glasgow's Sam Bayliss, "The change was realized just by interchanging single atoms on the host molecules, one of the smallest changes you could get, and it gave rise to the five-fold enhancement in coherence time."

Full Article

A scanning electron microscopy image of a breast cancer cell being attacked by T cells. Researchers 3D-Bioprint, Treat Breast Cancer Tumors
Penn State News
Adrienne Berard
October 18, 2022

Pennsylvania State University (Penn State) researchers have three-dimensionally (3D)-printed, then treated, breast cancer tumors. The researchers used a method called aspiration-assisted bioprinting to pinpoint tumors and generate tissue, which they fabricated into a multi-scale vascularized model. The team validated the model's accuracy by treating it with doxorubicin-based chemotherapy, then with cell-based immunotherapy. Penn State's Ibrahim Ozbolat said the bioprinted tumor model "serves as a clinical test platform to safely and accurately evaluate experimental therapies," as well as "a research platform for immunologists and biologists to understand how the tumor grows, how it interacts with human cells, and how it metastasizes and spreads in the body."

Full Article

Fiber-optic cables. Record 1.84 Petabit/s Data Transfer with Photonic Chip, Fiber-Optic Cable
Tom's Hardware
Mark Tyson
October 20, 2022

Researchers at the Technical University of Denmark have executed 1.84 petabits/second data transfers over 7.9 kilometers (4.9 miles) using a photonic chip linked via fiber-optic cable. This compares to the approximately 1 petabit/second Internet bandwidth used by the global population. The researchers split the data stream into 37 lines, and routed each line down a different optical thread in the cable. Each line was segmented into 223 data chunks corresponding with zones of the optical spectrum, generating a "frequency comb" in which data was transmitted in different colors simultaneously without disrupting other streams. This supported a "massively parallel space-and-wavelength multiplexed data transmission" system that vastly boosted the fiber-optic cable's potential data throughput.

Full Article

A screen capture of the Discmaster website. Lost Something? Search 91.7 Million Files from the ’80s, ’90s, 2000s
Ars Technica
Benj Edwards
October 18, 2022

A new website allows users to sift through 91.7 million computer files from CD-ROM releases and floppy discs dating back to the 1980s. Hosted by tech archivist Jason Scott, the Discmaster site is the work of a group of anonymous programmers and features images, text documents, music, games, shareware, videos, and more from the Internet Archive. Discmaster allows users to search by file type, format, source, file size, file date, and other criteria. Said Scott, "The value proposition is the value proposition of any freely accessible research database." Much of the file format conversion is performed on the back end, to make the vintage files more accessible.

Full Article

Computer scientist Juan Gilbert. Computer Science Has Racism Problem: Researchers Want to Fix It
Melba Newsome
October 19, 2022

The University of Florida (UF)'s Juan Gilbert is one of many researchers striving to address racism in computer science (CS). As chair of UF's Department of Computer & Information Science & Engineering, Gilbert fosters a more supportive and inclusive CS culture for Black students; he credits this for helping UF produce more Black CS Ph.D. graduates, and for more Black CS faculty, than any other predominantly white U.S. educational institution. In 2006, Gilbert founded the African American Researchers in Computing Sciences program to support Black Ph.D.’s in tenure-track faculty and research scientist positions in the CS-education pipeline. Duke University's Shaundra Daily sees a lack of diversity in CS reflected in computing algorithms, which can disproportionately impact marginalized people. Freeman Hrabowski, former president of the University of Maryland, Baltimore County, says addressing this problem requires academics and stakeholders to acknowledge and confront bias across the board.

Full Article

A drone image of salmon spawning. Drones Show Potential to Improve Salmon Nest Counts
Washington State University
Sara Zaske
October 19, 2022

A study by Washington State University (WSU) researchers found that drone imagery could be more effective than ground-level observations in estimating the number of rocky hollows, or "redds," made by salmon as a place to lay their eggs. Although drone imagery identified double the number of potential redds during spawning season on Washington’s Wenatchee River, whether each is a true redd can be determined only by disassembling the nest. WSU's Daniel Auerbach said, "While ground redd counts are less intrusive than other counts, drones are even less invasive, and we can use these images for many things," such as tracking habitat changes. WSU's Alex Fremier noted that high-resolution drone images "could be more cost-effective” than ground-level observations, and can help determine whether a restoration project actually helped build each salmon habitat.

Full Article

Interpol headquarters in Lyon, France. Interpol Setting Up Its Own Metaverse to Police the Virtual World
Oceane Duboust
October 20, 2022

Interpol has introduced what it describes as "the first-ever metaverse specifically designed for law enforcement worldwide." The global police organization said the tool will allow police officers the world over to interact via avatars, and to take "immersive training courses in forensic investigation and other policing capabilities." According to Interpol’s latest Global Crime Trend report, "As transnational crimes and criminal networks evolve at an unprecedented pace, law enforcement are often left playing 'catch up,' with tools, processes, and procedures that may be outdated or inefficient." Interpol said its metaverse is designed to police crimes in the virtual world, which can be associated with unique legal challenges.

Full Article

A man in a suit has a concerned look on his face while looking at a computer monitor. AI Tools Fail to Reduce Recruitment Bias
BBC News
Chris Vallance
October 14, 2022

A study by researchers at the U.K.'s University of Cambridge found that diversity and bias in hiring are not improved by the use of artificial intelligence technology. Said Cambridge's Kerry Mackereth, "These tools can't be trained to only identify job-related characteristics and strip out gender and race from the hiring process, because the kinds of attributes we think are essential for being a good employee are inherently bound up with gender and race." The researchers developed a simplified AI recruitment tool that assigns personality scores to candidates' photos based on agreeableness, extroversion, openness, conscientiousness, and neuroticism. However, they found that several irrelevant variables, such as changes in contrast, brightness, and saturation, skewed the ratings of their tool.

Full Article
Introduction to NVIDIA Modulus: A Physics-ML Framework for Research
ACM Online Books and Courses

Association for Computing Machinery

1601 Broadway, 10th Floor
New York, NY 10019-7434

ACM Media Sales

If you are interested in advertising in ACM TechNews or other ACM publications, please contact ACM Media Sales or (212) 626-0686, or visit ACM Media for more information.

To submit feedback about ACM TechNews, contact: [email protected]