Association for Computing Machinery
Welcome to the April 22, 2009 edition of ACM TechNews, providing timely information for IT professionals three times a week.


Robots Are Narrowing the Gap With Humans
McClatchy Newspapers (04/20/09) Boyd, Richard S.

At the recent RoboBusiness conference in Boston, companies demonstrated several advanced robots, including a robot firefighter, gardener, receptionist, tour guide, and security guard. A housekeeping robot developed in Japan can move chairs, sweep the floor, load dirty dishes into a dishwasher, and put clothes in a washing machine. Intel has developed a self-controlled mobile robot called Herb, short for Home Exploring Robotic Butler, that can recognize faces and execute commands such as "clean up this mess," says Intel's Justin Rattner. Last year, Rattner gave some credibility to the effort to make machines as intelligent as people, saying the industry has made significantly more progress during the past 40 years than people once believed possible. He said it is reasonable to believe that machines could overtake people in the ability to reason in the not-so-distant future. Nevertheless, experts say that programming a robot to perform tasks such as washing dishes is hard enough, and creating a truly intelligent robot is still a staggeringly difficult challenge. "One day we will create a human-level artificial intelligence,'' says Massachusetts Institute of Technology professor Rodney Brooks. "But how and when we will get there--and what will happen after we do--are now the subjects of fierce debate.'' Technology evangelist Ray Kurzweil predicts that robots will match human intelligence in 20 years, a point in time he calls the singularity. Stanford University's Paul Saffo says robot intelligence is making steady progress. "The truly interesting question is what happens after we have truly intelligent robots,'' Saffo says. "If we're very lucky, they'll treat us as pets. If not, they'll treat us as food.''

A Tool to Make More of Many Cores
Technology Review (04/21/09) Cringley, Robert X.

Later this year Intel will release a research project code-named Ct, short for C for Throughput, that will automatically make standard C and C++ compilers work with many-core processors--those with 16 to hundreds of cores. Ct will be part of Intel's Parallel Studio software-development toolkit and will enable programmers to employ all of the cores in many-core processors, ensuring they are running at full efficiency. Traditionally, creating parallel software has required determining which parts of the code were most easily adapted to parallel processors and isolating those parts in the module. However, isolating and applying parallel code would have to be done from scratch for each new processor family or with any significant increase in the number of cores. Intel says that Ct automates this process, and optimizing for many-core processors will not even require a recompile. Ct also is backward compatible with software written for Intel's x86 architecture. Ct will work with Intel's new graphics chip, code-named Larrabee, due next year. "Ct is a good match for Larrabee," says University of Illinois professor Marc Snir. "We have thought of Ct as something that is much more attractive than CUDA or OpenCL for developing data-parallel code."

Cheap and Noisy Chips Could Improve Climate Predictions
New Scientist (04/17/09) Inman, Mason

Researchers at the European Centre for Medium-Range Weather Forecasts say that running simulations on inexpensive computer chips that create results tainted with random noise could improve climate models. Researchers trying to make climate models more realistic are limited by the processing power of the supercomputers running the models, says researcher Tim Palmer. However, adding a degree of randomness to a particular model and running it several times could be an inexpensive way to improve the realism of climate modeling, Palmer says. If multiple, slightly different simulations of the same model create the same result, it indicates the strength of that prediction. This technique has already been successfully demonstrated in the prediction of weather during a period of several weeks. Intentionally generating randomness for these models would consume a significant amount of computing power, but using cheap computer chips that naturally create random noise due to how electrons move through them could provide an inexpensive and processor-efficient solution. "It's very speculative," Palmer says. "But if it can be made to work, it would make much more efficient use of power."

Soft Hardware for a Flexible Chip
ICT Results (04/21/09)

The European Union-funded MORPHEUS project is developing reconfigurable hardware by merging a microprocessor with reconfigurable units embedded in the same component. As specialized embedded systems become increasingly common, the technology is having difficulty keeping up with the necessary computing power. High-performance applications demanded by digital video processing, telecoms, and military applications, for example, are putting pressure on the industry. "This kind of equipment needs high computing performance for signal processing and for making decisions," says Philippe Bonnot of Thales Research and Technology, the coordinator of the MORPHEUS project. "But the solutions are not as efficient as we would like." The challenge is to create embedded systems that are efficient and flexible, which is normally not possible in the same chip. Reconfigurable hardware can be programmed to structure itself in a variety of ways, and when a new application is needed the hardware can be modified just like a piece of software to perform a different task. "The reconfigurable technology makes specific solutions possible," Bonnot says. "You can design exactly what you need so you are efficient, but it's reconfigurable so you can reuse the component for another application."

Tapia Conference Hits Record Attendance
HPC Wire (04/16/09)

Coalition to Diversify Computing (CDC) chair Pam Williams said this year's Richard Tapia Celebration of Diversity in Computing Conference was the best gathering so far. A record 400 people from academia, industry, and government attended the event, which took place April 1-4 in Portland, Oregon. There were 249 students and 73 faculty members from 101 universities and colleges present, and more than 20 companies and research institutions were involved in the conference, which is organized by CDC and sponsored by ACM. Industry, academic, and government supporters provided scholarships to 141 students in attendance. Ann Quiroz Gates, the former chair of the University of Texas at El Paso's Department of Computer Science and who now leads the Computing Alliance for Hispanic-Serving Institutions, received the Richard A. Tapia Achievement Award for Scientific Scholarship, Civic Science, and Diversifying Computing. Harvey Mudd College's "import antigravity" team won the robotics competition for the second straight year. The poster competition presentations also were impressive. "Appropriate for the Tapia conference, we had a very diverse group of posters, technically, from networking through weather modeling to education; geographically, from New York to the deep South; and demographically all sizes, shapes, and colors," says posters co-chair Brandeis Marshall from Purdue University.

Crowd Forms Against an Algorithm
New York Times (04/19/09) P. WK2; Rich, Motoko

An apparent cataloging error by that stripped thousands of books, many with homosexual themes, of their sales rankings and made them harder to locate sparked rampant accusations of bias and homophobia by online critics. The incident highlights the fallibility and controversial nature of the cataloging algorithms underpinning the Internet. "Whenever something like this happens, people immediately blame it on conspiracy," says University of Washington professor Ed Lazowska. "There are all kinds of ways for things to go wrong that are what I would call unintended consequences of either computer algorithms or human behavior." Cataloging systems that predated the Internet, such as the 19th century's Dewey Decimal System, were influenced by the cultural biases of their eras. "The ethical issue with algorithms and information systems generally is that they make choices about what information to use, or display or hide, and this makes them very powerful," observed Silicon Valley entrepreneur Mary Hodder in a blog post. "These choices are never made in a vacuum and reflect both the conscious and subconscious assumptions and ideas of their creators." David Weinberger at Harvard University's Berkman Center for Internet and Society says there is a general consensus that any single ordering of categories will directly reflect political and cultural prejudices, "and that's exactly where truth cannot lie."

Researchers Working on Memory to Replace DRAM, NAND
IDG News Service (04/21/09) Nystedt, Dan

Researchers at Taiwan's Industrial Technology Research Institute (ITRI) believe that resistive-random access memory (RRAM) could replace dynamic random access memory (DRAM) and NAND flash memory. They say RRAM could be ready for the embedded chip market within the next few years. Previous efforts to find a new type of memory to replace DRAM and NAND have focused on phase-change memory (PRAM), but ITRI researchers say PRAM is difficult to manufacture and they plan to wrap up their limited PRAM-related projects through the end of this year and phase them out. ITRI's Tsai Ming-jinn says RRAM is fast like DRAM, but can retain data when a device's power is turned off, which DRAM does not. ITRI has already produced 1Kbit prototype chips and successfully manufactured the chips on 8-inch wafers, but it will take years to create chips with enough storage capacity to make them of interest to the embedded chip market. "We're still in the early stage of development," Ming-jinn says. "Right now we cannot compete with DRAM on reliability."

Microsoft Cooking Up New Parallel Programming Language
eWeek (04/17/09) Taft, Darryl K.

Microsoft is developing a new language for parallel programming called Axum. Previously known as Maestro, Axum is an incubation project intended to help programmers handle parallel programming in the .NET environment. Microsoft has not committed to shipping Axum, but the language was recently demonstrated at the Lang.Net 2009 conference. Microsoft's Joshua Phillips says the Axum project aims to validate a safe and productive parallel programming model for the .NET framework. Axum builds on the architecture of the Web and the principles of isolation, agents, and message passing to increase application safety, responsiveness, scalability, and developer productivity, Phillips says. "Other advanced concepts we are exploring are data flow networks, asynchronous methods, and type annotations for taming side-effects," he says. "We currently have a working prototype with basic Visual Studio integration and a few demonstrations of working code." Phillips says Axum helps reduce complexity by eliminating implicit dependencies, providing a declarative model for dealing with state, and providing an application model that is inherently concurrent and responsive. He says Axum is actor-oriented and object-aware. "We're not talking about objects as a primary concept anymore; it's object-aware rather than object-oriented," Phillips says. "It's special-purpose, so we don't intend for Axum to be the general-purpose language that C# is."

'Instant On' Computing
National Science Foundation (04/17/09) Mixon, Bobbie

Researchers supported by the National Science Foundation have achieved a breakthrough in adding ferroelectric materials to silicon, without intervening reaction layers. The development could help other researchers in their effort to usher in instant access to computing, without users having to boot and reboot computer operating systems. Ferroelectric materials is the same low-power, high-efficiency electronic memory technology that smart cards use to instantly reveal and update stored information when waved before a reader. The team, led by Cornell University's Darrell Schlom, placed strontium titanate on silicon in a way in which the silicon would squeeze the strontium titanate into a ferroelectric state. "Several hybrid transistors have been proposed specifically with ferroelectrics in mind," Schlom says. "By creating a ferroelectric directly on silicon, we are bringing this possibility closer to realization." The development could lead to a new wave of memory devices that use less power and are faster and more convenient to use.

Study: People Manage Their Privacy on Facebook Naturally
Helsinki Institute for Information Technology (04/20/09) Noronen, Visa

Trust is a key factor in the way people manage their privacy when using social media tools, according to researchers at Finland's Helsinki Institute for Information Technology (HIIT). Although users cannot control what other people publish on Facebook and similar sites, they tend to provide only information about themselves that they want other people to see, and they avoid publishing negative information about other people, say researchers Airi Lampinen, Sakari Tamminen, and Antti Oulasvirta. People will limit their number of friends, and also will exchange private messages within more defined closed groups when they need to update their status, the researchers discovered during interviews with users of social media. "People protect their own privacy and other people's privacy instinctively, often almost without noticing," Lampinen says. "To support these activities, social networking sites need to provide users with easy-to-use privacy management that is interlinked with the overall use of the sites."

Dialect Detectives
MIT News (04/16/09) Ryan, Dorothy

Massachusetts Institute of Technology Lincoln Laboratory engineer Pedro Torres-Carrasquillo and his colleagues are developing a dialect identification system that could assist translators in identifying multiple variants of a spoken language. The researchers say the system could be used by law enforcement agencies to identify the origin of an intercepted phone call detailing a drug shipment, based on the dialect spoken and the region that dialect is from. Previous work at Lincoln Laboratory on dialect information focused on building models that mapped the audiowave frequencies of phonemes, the individual sounds of a spoken language. However, Torres-Carrasquillo has focused on lower-level acoustic systems that use the basic spectral similarities of small pieces of spoken utterances. "We are not looking for the types of data linguists deal with--larger units such as phonemes and words," Torres-Carrasquillo says. "We're looking at the statistical distributions of basic frequency spectra of small pieces of sounds." The researchers are building a model that classifies the training data and finds markers that discriminate the frequency chrematistics of the data. They are using pattern recognition and classification methods known as support vector machines and Gaussian Mixture models, which use models trained to emphasize the more distinctive tiny features seen in the frequency patterns of small pieces of the dialects in question.

Using Technology to Improve Traffic Flow and Road Safety
Universitat Politecnica de Catalunya (04/08/09)

The Polytechnic University of Catalonia (UPC) is collaborating on the MARTA project, an effort to develop technology that will make roads safer, ease traffic congestion, and lower emissions of contaminant gases. MARTA will incorporate sensors and interfaces in vehicles to collect traffic data, and information on traffic jams will be displayed on a screen or announced via an automated voice. On-board sensors also could monitor the status of brakes and control the driving speed and the distance between other vehicles on the road. Data would be shared between vehicles via the interfaces, and it would be transmitted to a mobility management center from a network of nodes along the roads. UPC's Research Group in Mathematical Programming, Logistics, and Simulation will design simulated traffic scenarios to test the new technologies. Meanwhile, UPC's Seat Chair of Innovation and Sustainable Development will help design and fit human-machine interfaces and run automated tests of the electronic systems. Along with national universities, 19 companies and 19 scientific centers are involved in project, which is expected to run through 2011.

Once Smartphones Become Truly Common, So Will the Viruses That Attack Them
Northeastern University News (04/14/09) Nyul, Renata

Northeastern University researchers say that smartphones will soon be targeted by viruses on a massive scale, but a study by the researchers could provide a way to negate these attacks. Northeastern University physicist and network scientist Albert-Laszlo Barabasi and fellow researchers tracked the spreading potential of Bluetooth and multimedia messaging service viruses, and predicted that these viruses will become a significant threat to smartphones that gain at least a 10 percent market share. The user base for smart, handheld devices is still small and fragmented, making a large virus outbreak impossible. However, Barabasi warns that once smartphones are more widely used and one of the operating systems increases its market share, the users of the system will be targeted by mobile viruses in only a matter of minutes. He says an outbreak on a smartphone could be worse than any outbreak on a traditional computer. Pu Wang, a Ph.D. candidate at Northeastern's Center for Complex Network Research, says understanding the basic spreading patterns of the viruses could help researchers find ways of minimizing their impact, estimate the realistic risk carried by mobile viruses, and develop measures to avoid the costly and damaging effects of outbreaks.

Abstract News © Copyright 2009 INFORMATION, INC.
Powered by Information, Inc.

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

Change your Email Address for TechNews (log into myACM)