For the most part, the study, which is authored by Titu Andreescu, Joseph A. Gallian, Jonathan M. Kane, and Janet E. Mertz, examines participation in the International Mathematical Olympiad (IMO), a grueling, nine-hour, six-problem essay examination. The study found that there have been numerous girls who have excelled in the IMO; however, the frequency with which girls of medal-winning ability are identified varies from country to country.

According to the study, countries that do well in the IMO have rigorous national mathematics curricula along with cultures and educational systems that value, encourage, and support students who excel in mathematics. Over the past 30-plus years, teams that have done well--Bulgaria, East Germany/Germany, and USSR/Russian--have sent 9, 10, and 13 different girls, respectively. At the same time, U.S. teams included just three girls. Additionally, the study revealed that U.S. participants are immigrants or children of immigrants from countries where education in mathematics is valued.

This pattern is reflected in other math competitions, including the USA Mathematical Olympiad and the Putnam Mathematical Competition for undergraduate students. The authors go on to say that, under certain conditions, females make up 11-24 percent of the children identified as having profound mathematical ability.

The study also examines faculty representation of women in five top U.S. research university mathematics departments. Just 20 percent of the women in these departments were born in the U.S. Of the 80 percent born elsewhere, many are immigrants from countries in which girls are frequently members of IMO teams.

-- Jonathan Erickson
jerickson@ddj.com

But according to David Peters, a professor of engineering who specializes in aircraft and helicopter engineering at Washington University in St. Louis (clearly a Cardinal's fan who still sees in his mind's eye Lou Brock stealing bases), it's all a matter of the player's center of gravity. More specifically, Peters says that dynamics equations can determine which slide gets you there more quickly, and that there are three important mathematical issues at play:

• Momentum. The mass of the body x how fast the player is moving. Peters explains: "There's angular momentum (mass movement of inertia times the rotational rate). If it's feet-first and you're starting to slide, your feet are going out from you and you're rotating clockwise; if it's head-first, as your hands go down, you're rotating counterclockwise."
• Newton's Law. Force is mass times acceleration.
• Finally, moments of inertia x your angular acceleration.

Right, but answer the question--who gets there faster?

"It turns out your center of gravity is where the momentum is," Peters says. "This is found half way from the tips of your fingers to the tips of your toes. In the headfirst slide, the center of gravity is lower than halfway between your feet and hands, so your feet don't get there as fast. It's faster head-first." Right, but try telling that to my son's Little League coach.

Peters goes on to say that "mathematically, you might think there's an advantage, but leaving your feet is actually a detriment because you're no longer pulsing (pumping your legs) and you start to decelerate. When you're running, you get your feet out in front of the center of gravity, so you're getting maybe three or four steps of an advantage."

So there you have it. Ricky Henderson might actually have known what he was talking about.

-- Jonathan Erickson
jerickson@ddj.com

And to what end? To analyze and manage more than 15 Petabytes (or 15 million gigabytes) of data generated annually by the Large Hadron Collider (LHC). Putting this in context, when the LHC starts running at full speed, it will produce enough data to fill six CDs per second, according to Michael Ernst, director of Brookhaven National Lab's Tier-1 Computing Center.

The LHC at CERN is the largest scientific instrument in the world. When it begins operations, it will produce roughly 15 Petabytes (15 million Gigabytes) of data annually, which thousands of scientists around the world will access and analyze. The mission of the Worldwide LHC Computing Grid project is to build and maintain a data storage and analysis infrastructure for the entire high-energy physics community that will use the LHC.

U.S. contributions to the Worldwide LHC Computing Grid are coordinated through the Open Science Grid (OSG), a national computing infrastructure for science. The OSG provides and supports the OSG Software Stack for download and use by administrators and users of the OSG. The OSG software stack relies on the Virtual Data Toolkit. The collaborative workspace (TWiki) pages provide detailed instructions on how to prepare a facility and/or resource and how to download and configure the OSG Software Stack in order to provide or access resources on the OSG.

-- Jonathan Erickson
jerickson@ddj.com

Of course, knowing that bugs can creep into hardware, CPU vendors do what they can to identify and simulate all possible configurations that might trigger bugs. Still, it is impossible to simulate every scenario, which is where the bugs come in. To keep bugs out, researchers at the University of Michigan are proposing a "virtual fence" that prevents a processor from operating in untested configurations. The technique keeps tabs of all tested configurations, and loads that information onto a "semantic guardian"--a monitor added to each processor that assumes all untested configurations are potential bugs. The semantic guardian lets the processor work only within its virtual fence by switching the processor into a slower safe mode when the processor encounters an untested configuration.

"If you consider all the possible configurations of the processor, only a tiny fraction of them is verified. But that tiny portion accounts for the configurations that occur 99.9 percent of the time," says Valeria Bertacco, who, along with Ilya Wagner, authored Engineering Trust with Semantic Guardians.

And, according to independent security researcher Kris Kaspersky, semantic guardians could also be a safeguard against exploiting hardware design bugs to gain control of other computers. "Semantic guardians would stop these security attackers dead in their tracks, since the processor would no longer be able to execute the buggy configurations that they were planning to exploit," explains Ilya Wagner.

-- Jonathan Erickson
jerickson@ddj.com

Adding fuel to the stereotype game player fire, Dmitri Williams, an assistant professor at the University of Southern California, conducted his own survey of 7,000 Everquest II gamers, and guess what--he discovered they didn't fit the gamer stereotype either. In his paper Who Plays, How Much, and Why? Debunking the Stereotypical Gamer Profile, Williams reports that there are more players in their 30s than in their 20s (the average age was 31), and playing time increased with age. Moreover, while women were only 20 percent of players, they spend more time in the game than men. (According to Sony Online Entertainment, EverQuest II players average about 26 hours per week.)

For the first time in online game research history, Sony Entertainment actually gave Williams and his co-authors--Nick Yee of the Palo Alto Research Center (PARC) and Scott Caplan of the University of Delaware--access to the game data. Yee is a member of the PlayOn Group, a research group at the PARC that explores the social dynamics of Massively-Multiplayer Online Role-Playing Games (MMORPGs). His role involves parsing and analyzing large data sets from longitudinal in-game census snapshots.

Gamers who participated in the survey, which spanned two days, earned a "Greatstaff of the Sun Serpent," a keepsake created specifically for the survey, in exchange for participation. Everyone who logged in during the recruitment period was offered the same prize.

-- Jonathan Erickson
jerickson@ddj.com

If you hang around Henry's coffee house as much as I do, you hear stories like this all the time. But Avik Ghosh, an assistant professor at the University of Virginia, tops them all: "If we continue at our current pace of miniaturization, these devices will be as hot as the Sun in 10 to 20 years." Which is why Gosh and fellow professor Mircea Stan are re-examining the Second Law of Thermodynamics that says, all else being equal, heat will transfer from a hotter unit to a cooler one until both have approximately the same temperature.

From what Ghosh and Stan have determined, if we could build components that operate outside this thermal equilibrium, laptops wouldn't get so hot. And as a bonus, performance would increase and battery life extended by using power dissipated by other functions. Of course first, the researchers have to solve Maxwell's Demon, a theoretical concept that suggests that the energy flow from hot to cold could be disrupted if there were a way to control the transfer of energy between two units. Maxwell's Demon would let one component take the heat while another worked at a lower temperature.

Another concept they're examining is Brownian ratchets, which suggests that devices could be designed to convert non-equilibrium electrical activity into directed motion, allowing energy to be harvested from a heat source.

"These theories have been looked at from a physics perspective for years, but not from the perspective of electrical engineering," says Stan. "So that's where we are trying to break some ground."

And if Ghosh and Stan are successful, I can already hear the crowd at Henry's: "My laptop runs so cool that I don't need air conditioning" or "So what? My laptop runs so cool...."

-- Jonathan Erickson
jerickson@ddj.com

Of course, this all could be avoided if wireless providers would identify--and compensate for--dead zones before building networks. However, this isn't cheap or easy, which is why WiFi providers build networks first, then plug holes later.

Apparently I'm not alone in my frustration, and I'm glad that Edward Knightly, a professor in electrical and computer engineering at Rice University, is frustrated by dead zones, too. But unlike me, Ed is smart enough to do something about it, as described in the paper Assessment of Urban-Scale Wireless Networks with a Small Number of Measurements that he co-authored with HP Labs research scientist Ram Swaminathan and Rice grad student Joshua Robinson.

The technique they came up with to tackle dead zones uses a small number of measurements to predict how well a wireless transmitter will cover a particular portion of a neighborhood. The only information required is basic topography, street locations, and general information about land use. The measurement framework that they present accurately characterizes the metric regions of a mesh note using a small number of measurements. Their primary focus is on coverage (signal strength).

As proof-of-concept, the team tested the technique both on Google WiFi, the free wireless Internet service Google provides in Mountain View, Calif., and TFA-Wireless, an experimental network in Houston, Texas, designed and built by Rice and owned and operated by the nonprofit Technology For All. TFA-Wireless provides high-speed Internet access to more than 4000 users in Houston.

"In the real world there are many things that can interfere with signals and limit coverage," says Knightly. "Our goal was to efficiently characterize the performance of urban-scale deployments, and our techniques can be used to either guide network deployment or to assess whether a deployed network meets its performance requirements."

And in the real world, this is a step to solving a real problem.

-- Jonathan Erickson
jerickson@ddj.com

According to Salmi (who holds a Ph.D. in sleep Sleep Medicine) and Virtanen (a software developer), HappyWakeUp is a smart alarm clock installed on your mobile phone that goes off when you're almost awake. HappyWakeUp doesn't bother you if you're in deep or REM sleep--times when it is most difficult for you to wake up. Instead, the program analyzes the quality and cycles of your sleep to determine your optimal get-up-and-get-out-of-bed moment.

During the final few hours of sleep, your sleep is more fragmented with short periods of wakefulness or "arousals"--the optimal moments for you to wake up. HappyWakeUp detects those moments by monitoring the quality of your sleep, based on noises that the microphone of your mobile phone picks up. During the final 20 minutes before the ultimate alarm time, a "soft" alarm will ring, depending on your snoring patterns and/or your movements. If you're sleeping calmly, the soft alarm doesn't ring.

"The alarm signal during deep sleep is stressful and familiar to everyone, but with the smart alarm clock this is avoided," says Salmi, who is researcher at University of Helsinki in Finland.

Interestingly, the Finnish researchers say that the same recording and analysis techniques can also be used to monitor sleep problems such as restless legs syndrome, snoring, and sleep apnea. For the time being, HappyWakeUp is only available for Nokia S60 smart phones.

As for what you'd name a product like this, well, I'll have some suggestions after a good night's sleep.

-- Jonathan Erickson
jerickson@ddj.com

Not that Paul Kintner and Mark Psiaki are mean spirited. In real life, Kintner and Psiaki are professors at Cornell University, and it's been my experience that anyone associated with Cornell is okay. But that still doesn't explain why they'd have so much fun spoofing GPS receivers. Of course, I'm sure they'd tell you that it's not fun--it's serious research. Which is why their description of what they do is described in a paper with the title of Simulating Ionosphere-Induced Scintillation for Testing GPS Receiver Phase Tracking Loops, instead of something like "GPS Spoofing for Fun and Profit."

Along with fellow researchers Brent Ledvina and Todd Humphreys, the team has described how a "phony" receiver could be placed in the proximity of a navigation device, where it would track, modify, and retransmit the signals being transmitted from the GPS satellite constellation. Gradually, the "victim" navigation device takes the fake navigation signals for the real thing.

By demonstrating the vulnerability of receivers to spoofing, researchers at the GPS Lab believe they can help devise methods to guard against such attacks. Right, and maybe get a chuckle out of seeing me walking north when I think I'm going south.

-- Jonathan Erickson
jerickson@ddj.com

"The stereotype that gaming is a solitary, violent, anti-social activity just doesn’t hold up. The average teen plays all different kinds of games and generally plays them with friends and family both online and offline," said Amanda Lenhart, author of a report on the survey and a Senior Research Specialist with the Pew Internet and American Life Project, which conducted the survey. Interestingly, virtually every teenager polled reported playing games. As for the commonly held stereotype that teenage gamers are solitary, anti-social basement dwellers that can't socialize? Forget it, says Lenhart. For most kids surveyed, games are a social experience, where they get to interact with their friends.

One thing the survey focused on was the relationship between gaming and civic experiences. Does playing games make teens less interested in their communities and less likely to becoming good citizens? Again, the survey found that teens who take part in social interaction related to the game--commenting on websites, contributing to discussion boards, etc.--are engaged civically and politically with the world around them.

Next was the team of Michigan Tech researchers led by Roshan D'Souza, who credits video gamers for pushing GPU-based technologies that's being used in agent-based modeling of complex biological systems, such as the human immune response to a tuberculosis bacterium (TB). "With a $1,400 desktop, we can beat a computing cluster," says D'Souza. "We are effectively democratizing supercomputing and putting these powerful tools into the hands of any researcher. Every time I present this research, I make it a point to thank the millions of video gamers who have inadvertently made this possible."

The University of Michigan's Denise Kirschner developed the TB model and passed it on to D'Souza's team, which programmed it using a graphic processing unit (GPU). Agent-based modeling hasn't replaced test tubes, she says, but it is providing a powerful new tool for medical research. Computer models provide a big advantage. "You can create a mouse that's missing a gene and see how important that gene is," says Kirschner. "But with agent-based modeling, we can knock out two or three genes at once." She adds that agent-based modeling lets researchers do something other methodologies can't--test the human response to serious insults, such as injury and infection.

"GPUs are very difficult to program. It is completely different from regular programming," said D'Souza, who deflects credit to the students. "All of this work was done by CS undergrads, and they are all from Michigan Tech. I've had phenomenal success with these guys--you can't put a price tag on it."

-- Jonathan Erickson
jerickson@ddj.com

Big deal. If you want to see a real 2-wheel man-and-machine test of skill and endurance, just watch me zip by on my motor scooter, heading for Henry's coffee shop for my mid-afternoon caffine fix. Desert, smesert. I weave in and out of pot holes, mud puddles, and dead squirrels at a breakneck speed. Breakneck for scooters, anyway.

What I don't have, of course, is a ARMOR X10 Rugged Tablet PC with an Intel SATA II (Serial Advanced Technology Attachment) solid-state drive mounted on my motor scooter. In the Baja race, an Intel team mounted the ruggedized PC on a Honda CRF450X motorcycle and left it running for the 11-hour duration of the grueling race. The test was proposed by Intel's Troy Winslow, an accomplished "weekend" motorcycle racer. Troy explains that "In this particular test we were doing a proof-of-concept for withstanding shock and vibration. We wanted to show that the unit was running throughout the race and could withstand severe shock and vibration. In our view, it was mission accomplished."

Except for the ruggedized chassis and housing, the ARMOR X10 is a fairly standard machine: 1.2-Ghz dual-core processor, 512 MB RAM, 16-GB solid-state hard drive, 10.4-inch LCD, touchscreen display, 4-6 hours of battery life, and it weights just over 5 pounds. While the ARMOR X10 sports a 16-GB solid-start hard drive, Intel does provide 32GB and 64GB versions of the X25-E Extreme SATA solid-state drive in the standard 2.5-inch hard-drive form factor. It goes 2 million hours mean time before failure and has a sustained sequential read of up to 250 MB/s and sustained sequential write of up to 170 MB/s. If you haven't been able to check out the emerging class of solid-state drives, do so. They're quiet, fast, and forget about cooling requirements.

And if DRS and Intel want to put an ARMOR X10 through a real acid test, let's strap one on my scooter and see how long it lasts. Now that would be something to brag about.

-- Jonathan Erickson
jerickson@ddj.com

"I call it a cube now, because it's not just a chip anymore," says Eby Friedman, director of the project. "This is the way computing is going to have to be done in the future. When the chips are flush against each other, they can do things you could never do with a regular 2D chip." The trick, say Friedman and Vasilis Pavlidis (co-authors of Three-dimensional Integrated Circuit Design), is to design a 3D chip where all the layers interact like a single system. Each layer could be a different processor with a different function--converting MP3 files to audio or detecting light for a digital camera, for instance. Friedman says that the 3D chip is essentially an entire circuit board folded up into a tiny package, and used inside something like an iPod could be compacted to 1/10th their current size with 10x the speed. A presentation by Friedman on 3D IC design is available here.

The 3D chip, which was manufactured at MIT, has millions of holes drilled into the insulation that separates the layers in order to allow for the myriad vertical connections between transistors in different layers.

-- Jonathan Erickson
jerickson@ddj.com

As it turns out, "Digital Fusion" was the theme of the CGSociety (that's short for "Computer Graphics Society") and NVIDIA's third annual digital art competition. And I have to say, in looking at the entries, the "wow!" factor was working overtime, which was the ideal considering that the competition called for "computer-generated imagery of 'awe-inspiring fusions' of 2D and 3D designs."

The first-place winner of the "Digital Fusion" competition was Fusion by Heiko Klug, second place Cock and the Centipede by Jason Lynch, and third place When Tlalli Calls You by Carlos Ortega Elizalde. Additional images of the winners, along with commentary about the winning entries, is available here.

Judges of the competition included Pascal Blanche, Ubisoft; Lorne Lanning, Oddworld Inhabitants; Stephan Martinere, Midway Games; Steven Stahlberg, Androidblues; Mark Snoswell, CGSociety; and David Wright, NVIDIA. They obviously know more about Digital Fusion than I do (I'm still dealing with the "how did they do that?" factor), but I bet I can out arc weld the bunch of them.

-- Jonathan Erickson
jerickson@ddj.com

In an effort to lighten the network load while still accommodating performance demands, researchers are working on P4P, a network architecture that optimizes traffic within individual ISPs, thereby reducing the volume of data traversing the ISP's infrastructure and creating a more manageable flow of data. According to initial tests, network loads can be reduced by a factor of five or more without compromising network performance, while at the same time, increasing speeds by about 20 percent.

Pushing the move to P4P is the DCIA P4P Working Group (P4PWG), which consists of ISPs, P2P software distributors, and researchers like the University of Washington's Arvind Krishnamurthy and Yale University's Richard Yang who have co-authored the paper P4P: Provider Portal for Applications. Other members of the Working Group include AT&T, BitTorrent, Cisco Systems, Pando Networks, Telefonica Group, and Joost, among others.

P4P can operate in multiple modes, ranging from ISPs who reveal their network status so that P2P applications can avoid hot-spots, to systems where markets and providers interact freely to create the most efficient information and cost flow. In a field test conducted using Pando software, P4P reduced inter-ISP traffic by an average of 34 percent, and increased delivery speeds to end users by up to 235 percent across U.S. networks and up to 898 percent across international networks.

All right! Can't wait to grab some music from the Trailer Park Troubadours at those speeds.

-- Jonathan Erickson
jerickson@ddj.com

Take the 60-foot long bus traveling on East 14th Street in San Leandro, California, a week or so ago. As far as buses go, it wasn't anything special--except that it didn't have a driver. Instead, it was traveling down the street via a magnetic guidance system, in which a series of magnets on the pavement were talking to special sensors and processors on the bus to control steering. Okay, a human did take care of braking and acceleration, but the steering was completely automated, allowing the bus to pull into stops to within a lateral accuracy of one centimeter. Still, in other tests the system has been used for full vehicle control--including braking and accelerating--creating a true "auto-pilot" system for the bus. The "Automated Bus Guidance System" project is funded by the California Department of Transportation and conducted by programs at the nearby University of California, Berkeley. According to engineers, the system allows for a bus to safely follow closely behind another in all kinds of weather and road conditions.

Not to be outdone, researchers at Carnegie Mellon University have teamed up with engineers at Caterpillar to develop autonomous versions of large haul trucks used in mining operations. The CMU Robotics Institute will be developing self-driving versions of Caterpillar's two largest haul trucks, each with payload capacities of 240 tons or more. Plans call for autonomous trucks to be integrated into some mine sites by 2010. The CMU team will be adapting perception, planning, and autonomous software architectures it originally developed for a DARPA autonomous vehicle program and the DARPA Urban Challenge robot race. Caterpillar was a major sponsor of the CMU Tartan Racing team that won last year's $2 million Urban Challenge.

As for me, well, little wheels keep on turnin'.

-- Jonathan Erickson
jerickson@ddj.com