Welcome to the January 30, 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).
HEADLINES AT A GLANCE
Optimizing Code
MIT News (01/30/17) Larry Hardesty
Researchers at the Massachusetts Institute of Technology (MIT) next week will present a modified version of a popular open source compiler that optimizes prior to adding the code needed for parallel execution at the ACM Symposium on Principles and Practice of Parallel Programming (PPoPP 2017) in Austin, TX. The compiler "now optimizes parallel code better than any commercial or open source compiler, and it also compiles where some of these other compilers don't," says MIT professor Charles E. Leiserson, who received the ACM Paris Kanellakis Theory and Practice Award for 2013 and the ACM-IEEE Computer Society Ken Kennedy Award for 2014. The enhancement stems from optimization approaches that already existed in the modified compiler. The main advancement is an intermediate representation using a fork-join model of parallelism, with the compiler's front end customized to a fork-join language called Cilk, which adds only two commands to the C programming language: the fork-initiating "spawn" and the join-initiating "sync." Cilk-written programs must explicitly tell the runtime when to check on the progress of computations and rebalance cores' assignments, with these invocations tracked by the compiler. The MIT team's intermediate representation adds three commands--detach, reattach, and sync--to a compiler's middle end. The reattach command specifies the recombination of parallel tasks' results, making fork-join code resemble serial code so many of a serial compiler's optimization algorithms will work on it without alteration.
Why More and More Torontonians Are Shelling Out $10K for Coding Crash Courses
CBC News (Canada) (01/30/17) Kate McGillivray
Boot camps in computer coding and Web design are growing increasingly popular in Toronto, Canada, offering students an attractive alternative or supplement to traditional university programs. These schools, which charge between $8,500 and $10,500 for full-time programs, provide intensive crash courses in computer science and do not require applicants to have any prior experience with coding. Jeremy Shaki, co-founder of Lighthouse Labs, says his Web development boot camps boast a 95-percent employment rate within 90 days of graduation. Hacker You, the first boot camp of its kind in Toronto, started with 30 part-time students in 2012 and expects to graduate nearly 1,000 students in its full and part-time programs in 2017. "About 40 percent [of students] have taken some kind of post-secondary education, primarily focused on softer skills, let's say history or social science," says Bitmaker CEO Andrew Mawer. "They are finding that they don't have a lot of tangible skills that employers want." Mawer says coding boot camps are a legitimate alternative to other traditional education options, while other camp leaders believe their courses complement four-year programs. Several boot camps in Toronto are registered as private career colleges, and others are built out of a university and offer the option of using boot camp experience as credits toward a master's degree.
New App Facilitates Mobility and Parking for People With Disabilities and Avoids Fraud
RUVID Association (01/27/17)
The European Union's SIMON project is working to solve problems associated with helping people with disabilities access public transportation and private vehicle transport. The project plans to respond to these challenges by offering a complete integration of technological solutions that facilitate accessible navigation, mobility information, and the management of access rights for parking badge holders. The solutions already have been tested on a large scale in Spain, Portugal, the U.K., and Italy. One of SIMON's most promising programs involves the combination of a mobile application and a new smart card model. The system provides users with information on accessibility and parking in real time so they are able to validate themselves as legitimate users of parking spaces. It also enables users to plan accessible routes using multimodal transport. The solutions proposed by SIMON also include applications and services for public authorities, public transport operators, and parking service managers. Cities that implement the SIMON system will manage the use of public parking spaces, receiving real-time information on the use of reserved parking spots, reducing fraud, and enabling inclusion policies to promote the sustainable use of all transport modes.
Smart City Transport Systems
A*STAR Research (01/25/17)
Researchers at the A*STAR Institute of High Performance Computing in Singapore have developed a machine-learning program to accurately recreate and predict public transport use based on the distribution of land use and amenities in Singapore. The researchers collected data from the city's smartcard system on people tapping in and out of individual bus and subway stations over a period of a week, totaling more than 20 million journeys. The smartcard data then was combined with city-wide information on how land was being used and high-resolution maps that identified individual amenities within a set radius of each station. The researchers tested three machine-learning models to find one that accurately reproduced, and then predicted, transport ridership across the city. "We found that a decision tree model performed best, with good accuracy, computational efficiency, and an easy-to-follow user display," says A*STAR researcher Christopher Monterola. The results suggest an increase in amenities of up to 55 percent across the city would increase ridership. The researchers found high-resolution amenity data is a much stronger predictor of ridership than general land-use details.
New High-Performance Computing Cluster at the Albert Einstein Institute in Potsdam
Max Planck Institute for Gravitational Physics (01/24/17) Benjamin Knispel
The new Minerva supercomputer at the Max Planck Institute for Gravitation Physics (Albert Einstein Institute) in Germany will help scientists compute gravitational waveforms and other complex simulations. To identify faint signals in the gravitational wave detectors' background noise and infer information about astrophysical and cosmological properties, scientists must calculate the mergers of many different binary systems with different combinations of mass ratios and individual spins. "Such calculations need a lot of compute power and are very time-consuming," says Max Planck Institute director Alessandra Buonanno. "The simulation of the first gravitational wave measured by [the Laser Interferometer Gravitational-Wave Observatory] lasted three weeks--on our previous supercomputer Datura. Minerva is significantly faster and so we can now react even quicker to new detections and can calculate more signals." With 9,504 compute cores, a 38-terabyte memory, and a peak performance of 302.4 teraflops, Minerva is more than six times as powerful as its predecessor. For the search for binary black hole mergers, the waveform models have been refined using the numerical and analytical solutions of Einstein's equations of general relativity. Researchers calibrated approximate analytical solutions with precise numerical solutions, enabling the available computing power to be used more effectively.
Finding Credibility Clues on Twitter
Georgia Tech News Center (01/26/17) Jason Maderer
Researchers at the Georgia Institute of Technology (Georgia Tech) have built a language model that identifies words and phrases that lead to strong or weak perceived levels of credibility on Twitter. The researchers scanned 66 million tweets linked to nearly 1,400 real-world events, and found the words of millions of people on social media have considerable information about event credibility. They focused on tweets surrounding events in 2014 and 2015, including the emergence of Ebola in West Africa, the Charlie Hebdo attack in Paris, and the death of Eric Garner in New York City. The researchers asked people to judge the posts on their credibility, and then fed the words into a model that split them into 15 different linguistic categories, each of which included positive and negative emotions, hedges and boosters, and anxiety. The team then examined the words to determine if the tweets were credible or not, and found they matched the humans' opinions about 68 percent of the time. "When combined with other signals, such as event topics or structural information, our linguistic result could be an important building block of an automated system," says Georgia Tech professor Eric Gilbert. The research will be presented in February at the ACM Conference on Computer-Supported Cooperative Work and Social Computing (CSCW 2017) in Portland, OR.
FPGAs Focal Point for Efficient Neural Network Inference
The Next Platform (01/26/2017) Nicole Hemsoth
Much recent research has focused on the hardware required for training neural networks and other machine-learning algorithms, but deep learning inference has its own unique challenges. A developing area in inference is the use of binarized neural networks (BNNs), which use binary weights and activations for all computations. BNNs supplant expensive floating point multiplications with a bitwise XNOR gate and left and right bit shifts. BNNs are well suited to reconfigurable logic devices, which contain fine-grained compute resources and can enable smaller, lower-power implementations. There are multiple ways BNNs can be implemented from a hardware perspective, and an investigation of field-programmable gate arrays (FPGAs) indicates they may deliver a low-precision solution. Research from Xilinx found high image classification rates, minimal latency, and high power efficiency when BNNs are mapped to FPGAs. FPGAs have a much higher peak performance for binary operations than floating-point networks, and the small memory footprint eliminates the off-chip memory bottleneck. The Xilinx researchers have developed a framework for a scalable BNN inference on an FPGA called FINN. "FINN-generated accelerators can perform millions of classifications per second with sub-microsecond latency, thereby making them ideal for supporting real-time embedded applications, such as augmented reality, autonomous driving, and robotics," the researchers say.
Vanishing Point: The Rise of the Invisible Computer
The Guardian (01/26/17) Tim Cross
Experts agree the steady advances of computer chip transistor shrinkage--faster chip speeds, greater efficiency, and less-expensive manufacturing--will soon reach their physical limits. This is not expected to stall the computer revolution, as scientists will have to tap new methods and materials to ensure continued upgrades. Better programming is one strategy, while another is redesigning chips that use more specialized hardware at the cost of general mathematical prowess. Other concepts seek to keep Moore's Law viable by stacking chip components in three dimensions, which could eliminate data retrieval delays by sandwiching layers of processing logic between layers of memory. On the more exotic front are quantum computers--but like three-dimensional chips, they only yield one-off improvements or apply only to certain kinds of operations. Still, these technologies will function well in data centers and help drive an even more significant trend: on-demand computing that requires a minimum of hardware. Users will increasingly use small, mobile devices and everyday appliances to harness vast computing resources maintained in remote warehouses, with apps such as Apple's voice-powered Siri digital assistant paving the way for this "Internet of things." However, this vision cannot be achieved without improving computers' energy efficiency, while innovations such as augmented reality could accelerate its realization.
Unhappy Developers Lead to Bad Code and Bad Processes
Network World (01/24/17) Steven Max Patterson
Researchers at the University of Stuttgart in Germany have examined the effects of software developers' state of happiness on performance and found unhappy developers adversely affect the development process and software products. The research could explain why some software companies treat their development teams with significant perks, such as ping-pong tables, foosball, and cappuccino machines. A clearer understanding of this correlation could lead to successful intervention and higher productivity. The findings were based on an analysis of the textual answers self-reported by developers to a questionnaire. Reduced cognitive performance is the biggest impact of unhappiness, as about 40 percent of developers self-reported that unhappiness impacted their work. In addition, unhappiness from negative situations produced mental unease such as low-self esteem, high anxiety, burnout, stress, and possibly depression. The researchers also found unhappiness leads to low motivation among developers, withdrawal from their work, and quitting jobs. Finally, unhappiness causes developers to take shortcuts in the development process and to deviate from the process or the agreed set of practices, which often leads to poor code quality.
Where Are the Software Engineering Jobs? In Cybersecurity
IEEE Spectrum (01/25/17) Tekla S. Perry
There are 1 million unfilled cybersecurity engineering jobs around the world, and that figure is expected to grow to 1.5 million by 2019, according to an Indeed.com report based on data from 10 countries covering the third quarter of 2014 to the third quarter of 2016. The report concluded the biggest mismatch between job openings and job seekers is in Israel, followed by the U.K. Software engineers looking for cybersecurity jobs in the U.S. and Canada have a little more competition, with only 66.7 and 68.1 job seekers, respectively, for every 100 open positions in those countries. The huge demand in Israel can be attributed to the country's position as a technology hub as well as its general emphasis on security. Employers are most interested in hiring network security specialists, especially in Israel, Ireland, the U.K., the U.S., and Germany. In the U.S., the specialty that is second-most in demand is application security. However, Indeed.com found there are more job seekers than openings for ethical hackers in some regions, especially in the U.S. and U.K.
I Can See Clearly Now
UNews (UT) (01/24/17) Vincent Horiuchi
University of Utah researchers have created "smart glasses" with liquid-based lenses that can automatically adjust focus on what a person is seeing, whether it is far away or nearby. The smart eyeglass lenses are made of glycerin, a thick colorless liquid enclosed by flexible rubber-like membranes in the front and back. The rear membrane in each lens is connected to a series of mechanical actuators that push the membrane back and forth like a transparent piston, changing the curvature of the liquid lens and the focal length between the lens and the eye. "The focal length of the glasses depends on the shape of the lens, so to change the optical power we actually have to change the membrane shape," says Utah professor Carlos Mastrangelo. He notes the lenses are placed in specialized eyeglass frames equipped with electronics and a battery to control and power the actuators. The bridge of the glasses has a distance meter that measures the distance from the glasses to an object via pulses of infrared light. When the user looks at an object, the meter instantly measures the distance and automatically adjusts the lenses. If the user sees another, closer object, the distance meter readjusts and tells the actuators to reshape the lens.
U.S. Intelligence Seeks a Universal Translator for Text Search in Any Language
Ars Technica (01/24/17) Sean Gallagher
The goal of the U.S. Intelligence Advanced Research Projects Activity's (IARPA) Machine Translation for English Retrieval of Information in Any Language (MATERIAL) program is to give researchers and analysts a tool to help them search for documents in their field of concern in any of the more than 7,000 languages spoken worldwide. IARPA is seeking an "'English-in, English-out' information retrieval system that, given a domain-sensitive English query, will retrieve relevant data from a large multilingual repository and display the retrieved information in English as query-biased summaries." Users would be able to search massive numbers of documents with a two-part query, first by listing the "domain" of the search in terms of what sort of information they are seeking, and then providing an English word or phrase describing the information sought. IARPA says another objective of MATERIAL "is to drastically decrease the time and data needed to field systems capable of fulfilling an English-in, English-out task." MATERIAL participants will receive access to a finite set of machine-translation and automatic-speech-recognition training data from multiple languages "to enable performers to learn how to quickly adapt their methods to a wide variety of materials in various genres and domains," according to the agency. "As the program progresses, performers will apply and adapt these methods in increasingly shortened time frames to new languages."
Abstract News © Copyright 2017 INFORMATION, INC.
To submit feedback about ACM TechNews, contact: [email protected]
|