Welcome to the March 31, 2017 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).
EU Ratchets Up the Race to Exascale Computing
HPC Wire
John Russell
March 29, 2017


Seven European Union (EU) countries--France, Germany, Italy, Luxembourg, Netherlands, Portugal, and Spain--last week signed a pact to establish a cooperation framework to advance an integrated world-class high-performance computing (HPC) infrastructure called EuroHPC. Developing exascale computers is a key aspect of this effort, which the European Commission (EC) sees as essential to maintaining the EU's competitiveness. "If we stay dependent on others for [HPC], then we risk getting technologically 'locked,' delayed, or deprived of strategic know-how," warns the EC's Andrus Ansip. "Europe needs integrated world-class capability in supercomputing to be ahead in the global race." The pact says the EuroHPC framework is intended to focus on "acquiring and deploying an integrated exascale supercomputing infrastructure...available across the EU for scientific communities as well as public and private partners, no matter where supercomputers are located." The signatories plan to have an implementation roadmap for the HPC infrastructure ready by year's end.

Full Article

A flash drive, SD and microSD cards next to a keyboard New Ultrafast Flexible and Transparent Memory Devices Could Herald a New Era of Electronics
University of Exeter
March 31, 2017


Researchers at the University of Exeter in the U.K. have developed new memory technology using a hybrid of graphene oxide and titanium oxide. The researchers say the new devices are low-cost and eco-friendly to produce, in addition to being well suited for use in flexible electronic devices. The devices also could have the potential to offer a less expensive and more adaptable alternative to flash memory, which currently is used in many common devices such as memory cards, graphics cards, and USB computer drives. The researchers say the new devices have the potential to revolutionize how data is stored and take flexible electronics to a new level of speed, efficiency, and power. "Using graphene oxide to produce memory devices has been reported before, but they were typically very large, slow, and aimed at the 'cheap and cheerful' end of the electronics goods market," notes University of Exeter professor David Wright.

Full Article

An “End stop and frisk” protest march Bias Test to Prevent Algorithms Discriminating Unfairly
New Scientist
Matt Reynolds
March 29, 2017


Researchers at the Alan Turing Institute in the U.K. are developing a framework to identify and eliminate algorithmic bias. A fair algorithm is one that makes the same decision about an individual regardless of demographic background. The researchers mapped out different variables in datasets and tested how they might skew decision-making processes. They applied this method to stop-and-frisk data from the New York City police department from 2014, modeling variables that influenced police officers' decisions to stop someone. The team analyzed the skin color and appearance of detained people, and they found police generally saw African-American and Hispanic men as more criminal than they did white men, a conclusion that could lead a machine-learning analysis to deduce that criminality is correlated with skin color. The researchers think this method could be applied to other organizations that are required to keep their processes free from discrimination.

Full Article

A loading image of a woman's face Faster Page Loads
MIT News
Larry Hardesty
March 28, 2017


Researchers at the Massachusetts Institute of Technology's Computer Science and Artificial Intelligence Laboratory have unveiled Flowtune, a new system for allocating bandwidth in datacenter networks. Tests showed the system maintained the same network throughput as those currently in use, while allocating bandwidth more fairly, completing the download of all of a page's components up to four times faster. Flowtune operators assign different values or "prices" to boosts in the transmission rates of data sent by different programs. For each pair of sending and receiving computers, Flowtune estimates the transmission rate that maximizes total "profit," or the difference between the value of higher transmission rates and the price of the requisite bandwidth across all the intervening connections. Maximizing profit changes demand across the links, so Flowtune continually recalculates prices as well as maximum profits, assigning the resulting transmission rates to the servers sending data across the network.

Full Article
Google Brain Wants Creative AI to Help Humans Make a 'New Kind of Art'
Technology Review
James Temple
March 29, 2017


Researchers at Google Brain are developing tools that match artists with deep-learning tools to develop novel artwork together. One platform, called Magenta, will enable people to produce completely new kinds of music and art. The researchers are continually trying to improve Magenta's algorithms for creating songs and producing art transfers from images. They say one critical challenge is developing better human interfaces for the technology. Although the researchers note they began with the equivalent of a command-line prompt, they want to get closer to the "naturalness" of a guitar stomp pedal. "I don't think that machines themselves just making art for art's sake is as interesting as you might think," says Google Brain researcher Douglas Eck. "The question to ask is, can machines help us make a new kind of art?"

Full Article

A fingerprint on a binary code background PolyU Develops Accurate Contactless 3D Fingerprint Identification System
Hong Kong Polytechnic University
March 30, 2017


Researchers at the Hong Kong Polytechnic University (PolyU) have created a three-dimensional (3D) contactless fingerprint identification system using new 3D fingerprint minutiae recovery and matching technology. PolyU professor Ajay Kumar's team has surmounted the constraints of contact-based two-dimensional (2D) biometric scans. Kumar's research represents 3D fingerprints as minutiae height and minutiae orientation in 3D spaces atop representations in 2D spaces, so the 3D system can more uniquely represent fingerprints with the additional data of two more measurements. The PolyU system employs a single digital camera that is paired with computer-controlled light-emitting diodes. Researchers can efficiently obtain high-frequency information in 3D fingerprints using advanced proprietary 3D fingerprint template-generation algorithms to retrieve 3D minutiae features. The system can match 3D fingerprints with about 97-percent accuracy while processing the data about two seconds faster than existing commercial 3D fingerprint-identification systems.

Full Article
Nanomagnets for Future Data Storage
Swiss Federal Institute of Technology in Zurich
Fabio Bergamin
March 30, 2017


Researchers at the Swiss Federal Institute of Technology in Zurich (ETH Zurich) have led an international team in the development of a technique for depositing single magnetizable atoms onto a surface, with implications for miniature data storage devices. ETH Zurich professor Christophe Coperet guided the development of a molecule with a dysprosium atom at its center, enclosed by a molecular scaffold functioning as a vehicle. The team also developed a method for depositing the molecules on the surface of silica nanoparticles and combining them by annealing at 400 degrees Celsius. The scaffold disintegrates and releases nanoparticles with dysprosium atoms, which can be magnetized and retain their magnetic information. The researchers now are attempting to stabilize magnetization longer and at higher temperatures, as well as fusing atoms to a flat surface. The nanoparticles can be stored at room temperature and reused, notes ETH Zurich's Florian Allouche.

Full Article
Clemson Scientists Receive $2.95M to Improve and Simplify Large-Scale Data Analysis
The Newsstand (SC)
Jim Melvin
March 30, 2017


The U.S. National Science Foundation has awarded researchers at Clemson University a $2.95-million grant for developing cyberinfrastructure to provide U.S. scientists with a more fluid and flexible large-scale data analysis system. Clemson professors Alex Feltus and Melissa Smith will lead the design of the Scientific Data Analysis at Scale (SciDAS) system. "We'll be processing scientific data at the same time that we're gluing together all the parts needed for a national cyberinfrastructure ecosystem," Feltus says. SciDAS will integrate access to multiple national cyberinfrastructure resources, petascale supercomputers, and various university resources. Scientists will leverage the distributed and scalable nature of both the data-sharing and the computer infrastructure to improve the execution of workflows and scientific productivity. "We intend to speed up the discovery process and complex end-to-end data analysis process through a tight coupling of science and cyberinfrastructure experts," Feltus says.

Full Article
How Graphene Could Cool Smartphone, Computer, and Other Electronics Chips
Rutgers Today
Todd B. Bates
March 27, 2017


Researchers at Rutgers University have discovered a way to use graphene to cool tiny chips. The researchers demonstrated a powerful and efficient cooling mechanism using graphene combined with a boron nitride crystal substrate. "We've achieved a power factor that is about two times higher than in previous thermoelectric coolers," says Rutgers professor Eva Y. Andrei. She notes if a piece of metal has one hot end and one cold end, the metal's atoms and electrons move quickly at the hot end and slowly at the cold end. The researchers applied voltage to the metal, sending a current from the hot end to the cold end. They say this process spurred the electrons to carry away the heat much more efficiently than via passive cooling techniques, such as little fans in computers.

Full Article

A doctor holding a clipboard and his patient sitting on a couch Psychologists Enlist Machine Learning to Help Diagnose Depression
Texas Advanced Computing Center
Aaron Dubrow
March 27, 2017


Researchers at the University of Texas at Austin (UT Austin) are using the Texas Advanced Computing Center's Stampede supercomputer to recognize patterns in neuroimaging data that are predictive for mental disorders such as depression. They are training a machine-learning algorithm that can spot common features among hundreds of patients using magnetic resonance imaging brain scans, genomics data, and other pertinent factors, to provide accurate predictions of risk for people with depression and anxiety. One recent study analyzed brain data from 52 treatment-seeking participants with depression and 45 healthy controls. "We feed in whole-brain data or a subset and predict disease classifications or any potential behavioral measure such as measures of negative information bias," says UT Austin professor David Schnyer. UT Austin professor Christopher Beevers notes, "one of the benefits of machine learning...is that machine learning...should generalize to new data."

Full Article
Can Robots Write Meaningful News?
Jonkoping University (Sweden)
March 27, 2017


Researchers at Jonkoping University's Media Management and Transformation Center in Sweden have launched the Digital Personalization of the News (DPer News) project to explore whether robots can enhance journalism. Project director Mart Ots says DPer News focuses on the question of "how can algorithms replace humans in repetitive professions? Journalism may not seem like a repetitive job, but when it comes to writing about finance and sports, it very well can be." The project's goal is to find creative techniques for robotization that can help the news industry produce more interesting news. "DPer News is about how we can make news stories that are not just cheap and convenient, but more meaningful and personal," notes Jonkoping professor Daved Barry. "It worries me that just because we can get robots to mine and condense data, that's all we'll do."

Full Article
A&M, NASA Partner to Design 'Robonaut'
The Battalion (TX)
Mikayla Andrade
March 27, 2017


Researchers at Texas A&M University (TAMU) are working with the U.S. National Aeronautics and Space Administration to develop a robot to assist astronauts in space. The researchers aim to create localization mapping algorithms for a robot, to be called Robonaut, that will help astronauts with simple tasks to save time and improve effectiveness. The ultimate goal for the project is to create a fully autonomous system that can navigate a space station and handle a wide range of tasks, says TAMU researcher Shu-hao Yeh. TAMU professor Dezhen Song notes although the Robonaut could be ready in only a few months, the project has many steps and the system can always be improved. The researchers are optimistic their project will impact the future of space exploration and society in a positive way. "If it functions according to our plan, it will make astronaut life easier," Song says.

Full Article
Cybersecurity Expert Studies Novel Tools to Thwart Attacks
CSUF News Center
March 27, 2017


California State University, Fullerton (CSUF) professor Mikhail Gofman says maintaining cybersecurity "requires constant work, commitment, and dedication," and he notes CSUF is researching new computer security tools. Among the projects Gofman and his students are involved in are some concentrating on next-generation multi-biometric authentication for mobile devices. Another solution his team is working on prevents sensitive information from reaching a computer's persistent storage. "It is naive to presume that a tool alone can stop a gamut of highly sophisticated cyberattacks aimed at influencing elections--especially targeted attacks orchestrated by state-actors designed to steal information of high importance and strategically disrupt the target's system," Gofman says. He also stresses effective election security solutions will demand "a comprehensive cybersecurity strategy that is as adaptive and persistent as attackers themselves. It must address not only the technical aspects, but also the human aspect--a single irresponsible decision by one individual can be sufficient."

Full Article
MIT Advances in CAD for Manufacturing
 
ACM Digital Library
 

Association for Computing Machinery

2 Penn Plaza, Suite 701
New York, NY 10121-0701
1-800-342-6626
(U.S./Canada)



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: technews@hq.acm.org