Entries with tag stanford university.

Researchers Demonstrate Accelerator-on-a-Chip

Scientists from the US Department of Energy’s SLAC National Accelerator Laboratory and Stanford University have created an advanced accelerator-on-a-chip technology that is smaller than a grain of rice. The glass chip can accelerate electrons at 300 million volts per meter, 10 times greater than conventional technology. The researchers’ goal is reaching speeds of 1 billion volts per meter. Over a distance of 100 feet, the technology can achieve acceleration comparable to that of the two-mile-long SLAC linear accelerator. Particle acceleration is primarily a tool used by physics researchers, but the new accelerator-on-a-chip – which is faster, smaller, and potentially a more affordable technology -- could be used in a wide variety of applications. This might include small, portable X-ray sources that could be used in field hospital settings by the military as well as for affordable medical imaging in hospitals and laboratories. The technology could also be used in for other types of scanning and imaging, such as in security applications. They say they must overcome some challenges before their technology is practical for routine use. They published their research in Nature. (SlashDot)(EurekAlert)(SLAC National Accelerator Laboratory)(Nature)

Stanford Researchers Build First Carbon-Nanotube Computer

Stanford University researchers have created the first working computer built exclusively using carbon-nanotube transistors. The computer can run a basic operating system, perform calculations, and switch between multiple processes running at the same time. The announcement is significant because researchers are looking for ways to replace silicon in transistors in order to create faster processors and higher-performing computers. The first nanotube transistor was developed in 1998. Researchers worldwide have been working to overcome problems inherent in growing carbon nanotubes and ensuring that the resulting device has no flaws that would impede its use. The Stanford researchers made the system on a single wafer with 197 dies. Each die contains five nanotube computers. Each computer consists of 178 carbon nanotube transistors. The device was made with standard chip-fabrication techniques and design tools, which would make the systems suitable for production in conventional facilities. Stanford professor Philip Wong said the system is a rudimentary first step in developing this type of computer. They published their research in Nature. (The Wall Street Journal)(Stanford News Service)(Nature -- 1) (Nature -- 2) 

Server Farms Can be More Sustainable

New research claims server farms can be made more sustainable through a combination of using clean energy, modern equipment and locating data farms in cooler climates. The report -- by Stanford University, Northwestern University and the Lawrence Berkeley National Laboratory researchers – claims these moves could make these facilities
88 percent more sustainable. Using current IT equipment alone, could reduce server farm greenhouse gas emissions by 80 percent, according to the researchers. Another 8 percent reduction could be possible if these facilities were in cooler locations, requiring less energy. Another significant problem, according to the study, are so-called zombie servers, hardware that stays on and uses energy, but is not being used for computational tasks. The findings were published earlier this summer in the journal Nature Climate Change. (The Guardian)(Nature Climate Change)

Stanford Team Demonstrates Nanotube Chip

Stanford University researchers have demonstrated a computer chip consisting of carbon nanotube-based transistors. The team—led by professor Philip Wong—created the chip with semiconducting carbon nanotubes. The resulting device is highly conductive and can be significantly smaller than silicon-based devices, which is increasingly important as device sizes continue to shrink. However, growing nanotubes with semiconducting properties is challenging. Today, most methods create batches of nanotubes in which as many as 30 percent are metallic rather than semiconducting. To address this, the scientists grouped the nanotubes in a way that self-corrects errors, allowing the device to continue working despite flawed materials. Rather than explain this process, they built a chip for analog-to-digital conversion using the carbon nanotube circuit. The scientists next plan to determine whether their technology could scale enough for use as a CPU in conventional computers. They demonstrated their technology at the recent IEEE International Solid-State Circuits Conference in San Francisco.(PhysOrg)(Technology Review) 

New Supercomputing Record Set

Scientists at the Stanford University-based Center for Turbulence Research, which is operated by both the school and NASA, set a new supercomputing record by using a million processing cores to model supersonic jet noise. They used the US Lawrence Livermore National Laboratory’s IBM Sequoia Blue Gene/Q system to solve the complex fluid-dynamics problem. Their work could not only help develop quieter aircraft engines, but also proves that million-core simulations are possible. (EurekAlert)(Stanford University) 

Stanford Using RFID, Online Gaming to Encourage Off-Peak Commuting

Stanford University professor Balaji Prabhakar is trying to use a mix of technology and incentives to reduce rush-hour traffic congestion. With a $3 million grant from the US Department of Transportation, he created Congestion and Parking Relief Incentives (CAPRI) to address commuting issues on campus. A participant receives an RFID-based ID tag for their vehicle’s windshield. Scanners at each main campus entrance detect the tags. For off-peak entry and departures, CAPRI awards credits that participants use to play an online game Prabhakar developed that offers cash prizes of between $2 and $50. Thus far, the program has awarded $31,000 in prizes. Prabhakar will extend the study this fall to reward drivers who park in less-used parking areas. The CAPRI system also links with social media, letting participants gain additional credits for enrolling friends. Prabhakar said they are considering other incentives that could, for example, target the student commuters and may add smartphone functions to the system. “So far, the CAPRI team hasn’t publicly released any specific data on how well the program has worked,” according to Ars Technica, “but some users have reported a dramatic drop in their own commuting times—as large as dropping from 25 minutes to 7 minutes.” Prabhakar has similar traffic-congestion-relief systems in Singapore and India. (Ars Technica)(Discover Magazine)(Stanford University)(CAPRI)

Computational Biologists Model Complete Organism

Stanford University researchers used information from more than 900 scientific papers to complete the world’s first computer model depicting the functions of an organism. The team described the various interactions that occur in the bacterium Mycoplasma genitalium and say their work is a step toward using CAD in bioengineering and medicine. It could also help scientists better understand cellular function and create new approaches for the diagnosis and treatment of disease. The bacterium modeled has only 525 genes, compared to E. coli, which has 4,288. Despite the organism’s small size, the scientists needed more than 1,900 parameters in 28 different modules to build their computational cell. They reported their work in the journal Cell. (Science Daily)(Stanford University)(Cell)

New Authentication Approach Uses Implicit Learning

Stanford University researchers have developed a new authentication approach based on a cognitive psychology concept called implicit learning. In this concept, subjects learn patterns without being aware they are doing so. Via a computer game, the system delivers a secret password to the user without the user consciously knowing what the password is. The system uses patterns in playing the game that create the password. In testing, participants reauthenticated the series of 30 keystrokes by re-playing the game. Organizations could use the approach in high-security applications such as systems providing access to a secured area of a nuclear or military facility. With their system, researchers say, users could not be coerced into revealing the password because they aren’t conscious of it. They will be presented their work on 8 August at the USENIX Security Symposium in Bellevue, Washington. (PhysOrg)(NBCNews.com)(New Scientist)

Showing 8 results.