Despite all the promise of computer components at the nanoscale and nanoscale devices, there is a downside: Experts say that the circuits and chips may be less reliable and more expensive to produce, leading to variability in behavior a device over its life.
The National Science Foundation (NSF) recently awarded a grant of U.S. $ 10 million from five years to help solve the problem. Expeditions The funding agency in the computer program funds projects that “promise significant advances in the computing frontier and great benefit to society,” has given the green light to a research project to rethink and improve the software function that can play in a new class of computing devices that are adaptable and highly efficient energy. Specifically, the research team is exploring “Variability-Aware Software for efficient computing with nanoscale devices,” the mission of developing computer systems that sense the nature and extent of changes in hardware circuits and expose their these variations to the compilers, operating systems and applications, adaptations in the software driving the whole stack.
Such computer systems for variability-would benefit the whole range of embedded, mobile, desktop and server class applications by dramatically reducing hardware design and test computer systems costs, while improving performance and efficiency energy, researchers say. Applications like search engines and medical imaging systems would also benefit, but the initial emphasis of the project team will be in wireless sensing, radio software and mobile platforms with plans to transfer these gains to other areas of forward motion.
“We envision a world in which the components of the system – led by proactive software – routinely monitor, predict and adapt to the variability of manufactured computer systems,” said Rajesh Gupta in a prepared speech on the grant the project. Garcia is the director of the expedition Variability and a professor of computer science and engineering at the University of California, San Diego Jacobs School of Engineering. “Change the way software interacts with hardware offers the best hope to perpetuate the fundamental achievements in computing performance at a lower cost of the last 40 years.”
As the transistors on the chips and components become smaller, semiconductor manufacturers are experiencing lower yields and greater variability, which means more components are being discarded because they do not comply with the schedule-, power-and specifications related to life. The program’s researchers argue that the trend towards the parts can not be expanded in capacity reliability or cost computing and paralyze the industries of information technology, if not addressed. A software-hardware interface fluid, the researchers maintain, will mitigate the variability of the systems manufactured and made more robust, reliable and responsive to changing operating conditions.