Association for Computing Machinery
Welcome to the September 16, 2013 edition of ACM TechNews, providing timely information for IT professionals three times a week.

Updated versions of the ACM TechNews mobile apps are available for Android phones and tablets (click here) and for iPhones (click here) and iPads (click here).


The Brogrammer Effect: Women Are a Small (and Shrinking) Share of Computer Workers
The Atlantic (09/13/13) Jordan Weissmann

Women account for 26 percent of science, technology, engineering, and math (STEM) workers, but the figure among computer workers has been dropping over the past 20 years, according to a recent U.S. Census Bureau report. Although women represented one-third of computer workers in 1990, they currently hold just 27 percent of jobs in this field. The gender disparity varies by computer specialty, with women representing 40.1 percent of database administrators and 37 percent of Web developers, but only 22.1 percent of software developers and 11.4 percent of computer network architects. The number of women as a percent of all computer science undergraduates reached its highest point in the 1980s, with women in the computer workforce peaking shortly thereafter. Women's low participation in the STEM workforce is partly responsible for the pay gap between the sexes, since STEM jobs generally have higher salaries. Moreover, because computer fields make up half of all STEM positions, according to the Census Bureau, the declining rate of women is particularly troubling.

Managing Multicore Memory
MIT News (09/13/13) Larry Hardesty

Massachusetts Institute of Technology (MIT) researchers have developed Jigsaw, a system that monitors the computations being performed by a multicore chip and manages cache memory accordingly. In experiments simulating the execution of hundreds of applications of 16- and 64-core chips, Jigsaw was able to accelerate execution by an average of 18 percent while reducing energy consumption by as much as 72 percent. Jigsaw monitors which cores are accessing which data most frequently and calculates the most efficient assignment of data to cache banks. Jigsaw also varies the amount of cache space allocated to each type of data, depending on how it is accessed, with the data that is reused frequently receiving more space than data that is accessed less often. In addition, by ignoring some scenarios that are extremely unlikely to arise in practice, the researchers developed an approximate optimization algorithm that runs efficiently even as the number of cores and the different types of data dramatically increases. MIT professor Daniel Sanchez notes that since the optimization is based on Jigsaw's observations of the chip's activity, "it's the optimal thing to do, assuming that the programs will behave in the next 20 milliseconds the way they did in the last 20 milliseconds."

Vote Early, Vote Often: Inside Norway's Pioneering Open Source E-Voting Trials
ZDNet (09/13/13) Stig Oyvann

Norway recently held its second e-voting pilot, following an initial trial that took place during the local government elections in 2011. In the second pilot, e-voting participation increased significantly compared to 2011, according to the Ministry of Local Government and Regional Development. Citizens in 12 municipalities could vote between Aug. 12 and Sept. 6 for the Aug. 9 election. In this year's election, 28 percent of all voters in the trial municipalities voted via the Internet, up from 16 percent in the 2011 pilot. To prove both the trustworthiness and transparency of the pilot, the source code for this year's e-voting system was put into the public domain, enabling citizens to download and study it. The electronic ballot used a public key mechanism for delivery. First, the vote was encrypted so that it could not be tied to the voter's identity. The vote then was digitally signed with the voter's public key, which kept it tamperproof. Each voter was given their own unique set of random codes for the different candidates in the election. Once a vote was cast, the code was texted back to the voter, enabling them to compare the code sent to their phone with the printed code on their voting card.

Girls-Only Coding Class Looks to Increase Female Tech Startup Presence (09/12/13) Michael Passingham

Female students at six of Britain's top universities will be able to take a free coding course this academic year designed to prepare them for the tech startup scene. Code First: Girls will be available at Oxford, Cambridge, the London School of Economics, University College London, Warwick, and Imperial College. The offering will be a condensed version of the program that was set up in 2012 as a nine-week, free, part-time course to get more women interested in coding, as well as contributing to technical discussions in high-tech businesses. The first run of the course was a success, and focused on key skills involving Sinatra and Rails and Ruby, as well as using communities such as GitHub. The course should make a big difference for young women who move on to startups, says Alice Bentinck, co-founder of startup accelerator scheme Entrepreneur First. "We find that more and more girls are coming to us with ideas, but they have no idea how to put them into practice," Bentinck notes.

CISE Issues Mid-Scale Infrastructure Solicitation
CCC Blog (09/10/13) Ann Drobnis

The U.S. National Science Foundation's (NSF) Directorate for Computer and Information Science and Engineering (CISE) recently announced the Mid-Scale Infrastructure (NSFCloud) initiative, which aims to support research infrastructure that enables the academic research community to develop and experiment with novel cloud architectures addressing emerging challenges, including real-time and high-confidence systems. The initiative "comprises a long-term and comprehensive program seeking infrastructure that will specifically enable the academic research community to (a) develop and experiment with novel cloud architectures, and (b) pursue architecturally-enabled novel applications of cloud computing," according to the NSF solicitation. The NSFCloud initiative also is designed to build on CISE's existing investments in infrastructure, particularly networking research infrastructure, that have demonstrated the value of developing and using shared infrastructure for accelerating research and education. "[We] seek proposals for research infrastructure that build upon existing investments as well as the recent rapid growth in cloud computing, and enable the academic research community to develop and experiment with novel cloud and cloud-like architectures that can support a diversity of innovative applications," the solicitation says.

MOOSE Enables 'Plug and Play' Simulations
HPC Wire (09/11/13) Tiffany Trader

Idaho National Laboratory (INL) researchers are developing the Multiphysics Object-Oriented Simulation Environment (MOOSE), a software framework for simulating the behavior of complex systems. The researchers aim to make simulation more accessible by making it easier to create simulation capabilities for complex mathematical models in various fields. The researchers say MOOSE opens up the advantages of simulation to experts so they can advance their science without also having to become computer scientists. "Something that would take five years with a team of 10 people can now be done in one year with three people," says INL's Derek Gaston. MOOSE was designed to be a general problem solver, capable of accommodating multiple mathematical models. "The user needs to know the governing equations for his or her field, and MOOSE solves them for you, meaning the scientist can focus on the science," says INL's Steve Hayes. In addition, MOOSE runs on personal workstations, so researchers can carry out powerful simulations without a supercomputer. The MOOSE ecosystem also has been expanded to provide tools for specific scientific disciplines, including applications for nuclear physics, geology, chemistry, and engineering.

New Magnetic Semiconductor Material Holds Promise for 'Spintronics'
NCSU News (09/10/13) Matt Shipman

A new compound that can be integrated into silicon chips potentially could be used to make spintronic devices, according to researchers at North Carolina State University (NCSU). The team created and synthesized strontium tin oxide as an epitaxial thin film on a silicon chip. The material is a single crystal and a dilute magnetic semiconductor, which means it could be used to create transistors that operate at room temperature based on magnetic fields, rather than electrical current. "We're talking about cool transistors for use in spintronics," says NCSU professor Jay Narayan. He notes the discovery that the material has magnetic semiconductor properties was a pleasant surprise. "There are other materials that are dilute magnetic semiconductors, but researchers have struggled to integrate those materials on a silicon substrate, which is essential for their use in multifunctional, smart devices," Narayan says.

'Tamper-Proof' Chips, With Some Work, Might Give Up Their Secrets
Technology Review (09/11/13) David Talbot

Technical University of Berlin researchers have discovered an attack involving costly equipment that can mill down the back of the silicon on chips and steal the data with microscopic probes. The research "is nice work that establishes that there is a new class of attacks that should be considered if invasive attacks are a concern," says Massachusetts Institute of Technology professor Srini Devadas. The researchers first used a polishing machine to gradually mill the back of the silicon until it was only 30 micrometers thick. Then they put the thinned chip under a scanning laser microscope fitted with an infrared camera and watched where key operations were happening. "We can see the heat emissions and know this is where it is running when the encryption algorithm starts to crunch numbers," says IO Active's Christopher Tarnovsky, who collaborated on the research. The researchers then used a focused ion-beam machine to dig tiny trenches to edit features on the chip, which made it possible to use minute probes that could essentially wiretap communications channels on the chip and access data.

Tough Robo-Challenge Casts Robots as Rescuers
New Scientist (09/11/13) Hal Hodson

The U.S. Defense Advanced Research Projects Agency (DARPA) Robotics Challenge is bringing together leading global robotics specialists in a competition to design a robot that can perform emergency-response duties during disasters. In December, the robots will compete in an obstacle course featuring eight challenges, such as traveling on uneven ground, getting into and driving a rescue vehicle, breaking down a wall, and shutting off valves. Boston Dynamics worked with Massachusetts Institute of Technology (MIT) researchers to create the Atlas humanoid robot. Robots excel at specific tasks, such as moving towards and picking up an object, says MIT's Russ Tedrake. "Where they fall down is planning what to do next. That's where the human comes in," Tedrake says. Atlas is monitored by a team of humans who review the robot's decisions and planned movements. A less humanoid contestant is the National Aeronautics and Space Administration Jet Propulsion Laboratory's entry called RoboSimian, which resembles a cross between an orangutan and a wolf spider. RoboSimian's sensing equipment is inside its body and it has four arms with seven joints each, providing more options for navigating an obstacle course. Carnegie Mellon University is creating a humanoid robot called CHIMP, with legs that can fold up and serve as treads for rough terrain.

A Swarm on Every Desktop: Robotics Experts Learn From Public
Rice University (09/09/13) Jane Boyd

Rice University researchers working in the Multi-Robot Systems Laboratory (MRSL) are refining their control algorithms for robotic swarms based upon data from five free online games. In the games, players use simple commands to move groups of robots through mazes and around obstacles. Each time a game is played, the website collects information about how the task was completed. The data will be used to help develop new control algorithms for robot swarms, says MRSL's Aaron Becker. "The data from these games will help us better understand how to use multi-robot systems with massive populations to perform coordinated, complex tasks," says Rice professor James McLurkin. In one experiment, a swarm of randomly scattered r-one robots were instructed to form a capital R. The robots were controlled with a one-button, 1980s-era videogame joystick that was only capable of telling the robots to either rotate or roll forward. In the experiment, a control algorithm directed each r-one robot in the swarm to a unique, pre-programmed end position. Although the researchers acknowledge that the current algorithm is slow, they say data from the online games will be used to design new control algorithms that are up to 200 times faster.

Wireless Network Detects Falls by the Elderly
University of Utah News (09/09/13) Aditi Risbud

New fall-detection technology developed by a team at the University of Utah would not require the elderly to wear any monitoring devices. Researchers Brad Mager and Neal Patwari have built a monitoring system using a two-level array of radio-frequency sensors placed around the perimeter of a room at two heights that correspond to someone standing or lying down. The sensors are similar to those used in home wireless networks. Each sensor in the array transmits to another, and anyone standing or falling inside the network would alter the path of the signals sent between each pair of sensors. The fall-detection system also can distinguish between a dangerous fall and someone simply lying down on the floor. "Ideally, the environment itself would be able to detect a fall and send an alert to a caregiver," Patwari says. By measuring the signal strength between each link in the network, an image is produced to display the approximate location of a person in the room with a resolution of about six inches. The radio tomography imaging technique uses the one-dimensional link measurements from the sensor network to construct a three-dimensional image.

Cyborg Astrobiologist Uses Phone-Cam to Search for Signs of Life
ScienceDaily (09/10/13)

Researchers from the Free University of Berlin, West Virginia University, the University of Malta, and the Center for Astrobiology in Madrid have developed a hybrid part-human, part-machine visual system that uses a mobile phone camera to search for evidence of past or present life in planetary analog sites on Earth. In the Cyborg Astrobiologist system, the user takes images of the surroundings using a mobile phone camera. The images are then sent via Bluetooth to a laptop, which processes the images to detect novel colors and textures and communicates back to the user the degree of similarity to previous images stored in its database. "We are now working to speed up the image-compression analysis and put the whole system onto a smartphone--and eventually onto a Mars rover," says Free University's Patrick McGuire. Tests of the Cyborg Astrobiologist system have been conducted at field sites with similarities to landscapes that are found on Mars. "The novelty detection also worked well, although there were some issues in differentiating between features that are similar in color but different in texture, like yellow lichen and sulphur-stained coalbeds," McGuire says.

Abstract News © Copyright 2013 INFORMATION, INC.
Powered by Information, Inc.

To submit feedback about ACM TechNews, contact: [email protected]
Current ACM Members: Unsubscribe/Change your email subscription by logging in at myACM.
Non-Members: Unsubscribe