Items tagged with supercomputers
There's a good old fashioned nerd fight brewing between the eggheads at IBM and Google. Don't worry, it's not physical—it's a war of words and theoretical concepts. What sparked the debate is Google's claim of achieving "quantum supremacy" with Sycamore, a system outfitted with a 53-qubit processor and how it performs compared to a classical supercomputer in a specific task. Google conceived an experiment to demonstrate just how capable Sycamore is, noting it can complete a certain benchmark test in 200 seconds, whereas "a state-of-the-art supercomputer would require approximately 10,000 years to perform the equivalent task." The spunky Mountain View firm details the experiment in a blog...
Read more...
The UK Research and Innovation (UKRI) center announced it has contracted with Cray to build a supercomputer based on AMD's latest generation EPYC "Rome" server processors. This will be the second generation of ARCHER (Advanced Research Computing High End Resource), with ARCHER2 being much more powerful than its predecessor. "ARCHER2 should be capable on average of over eleven times the science throughput of ARCHER, based on benchmarks which use five of the most heavily used codes on the current service," UKRI stated in an email. The original ARCHER supercomputer entered the scene in 2013 and resides at the University of Edinburgh. It is based on the Cray XC30, with Intel Xeon hardware inside....
Read more...
In an effort to give the U.S. a leg up when it comes to supercomputers, the Department of Energy announced its plans to build two GPU-powered supercomputers that will bring the world closer to exascale computing. The DOE is awarding $325 million to build “Summit” for the Oak Ridge National Laboratory in Tennessee and “Sierra” at the Lawrence Livermore National Laboratory in California while an additional $100 million will go into research for “extreme scale supercomputing” technology. The supercomputers are expected to be installed in 2017 using next-generation IBM POWER servers coupled with NVIDIA Tesla GPU accelerators and NVIDIA NVLink high-speed GPU interconnect...
Read more...
The Green500 ranks the most energy-efficient supercomputers in the world, and the latest list has an Intel-powered system at the top: “Beacon”, a supercomputer located at the National Institute for Computational Sciences (NICS), is the world’s most efficient supercomputer. Notably, Beacon uses Intel’s new Xeon Phi 5110p coprocessors(running its MIC architecture) along with Intel Xeon E5-2670 chips to drink just 44.89kW of power while offering almost two-and-a-half billion floating-point operations per second per watt (2.49 GFlops/Watt) and hits 112,200 gigaflops at peak performance. Intel has to be feeling pretty good about that ranking, because it firmly believes that...
Read more...
Researchers at MIT have developed chip architecture that runs on 70 percent less voltage than a conventional processor. The savings in power consumption would have profound effects on the battery life you could expect from portable devices. They may even require only movement and body heat to power chip-enabled sensors and communication devices. "It will extend the battery lifetime of portable devices in areas like medical electronics," said Anantha Chandrakasan, a professor of electrical engineering at MIT. "When you look at the digital processor, the fact is that we may be able to reduce the energy needed by 10 times."Better circuit design and batteries have already led to smaller, more-mobile...
Read more...
It looks like the trend of making more energy efficient PCs has spread from common laptops, desktops and servers, all the way up to supercomputers. The system is designed for defense and scientific research. "Called Maxwell, the computer has been built at the University of Edinburgh and uses field programmable gate arrays (FPGA) in place of conventional processors. Its Scottish developers believe Maxwell represents a new generation of compact and energy-efficient computers. Unlike ordinary general-purpose processors, FPGA chips can be programmed to perform very specific tasks. Once that programming is accomplished, FPGA chips can be much faster than performing the same tasks in software...
Read more...
If you've ever wondered what some of today's top supercomputers are up to, this article on CNN explains many of the more common applications. Some of them are used for tasks like forecasting weather, designing safer more fuel-efficient cars, mapping DNA, exploring the cosmos and even calculating the curvature of Pringles potato chips. "Supercomputers allow researchers to do in real time -- meaning days, weeks or months -- what could not be done during a lifetime with a single personal computer. "The amount of data that some of these supercomputers [produce] would be, maybe as much as 100,000 times more data than what you can put...
Read more...
