We’re still simply getting down to explore the utility of supercomputers. They’re serving to U.S. perceive a lot of regarding a number of humanity’s most complex challenges, and they’re still pretty unaccustomed the scene.
Supercomputers area unit ready to simulate terribly complicated environments, like weather and climate, astronomical events, explosions, and rocket engines. They’re conjointly terribly valuable for military uses and national defense, that is one huge reason why nations area unit willing to pour such a lot cash into their creation and maintenance.
NASA is incredibly fascinated by all this, as you may guess, that is why they’re operating to enhance their progressive Pleiades mainframe. they require your facilitate within the $55,000 High Performance quick Computing Challenge, a contest to enhance the software package that this mainframe runs on. the competition is over, however you’ll still keep reading to find out all regarding the Pleiades mainframe, one amongst the foremost technologically advanced items of apparatus on the earth.
The Pleiades mainframe is found at the NASA Ames center in geographic region, California. It’s a section of the NASA Advanced mainframe Division (NAS), and is that the most advanced and powerful resource within the the High-End Computing Capabilities Project (HECC), one amongst NASA’s initiatives to accelerate scientific progress. the opposite supercomputers at NAS area unit Electra, Endeavor, and Merope.
Supercomputers area unit an enormous deal, thus every one lands up obtaining its own name. This one is called once the Pleiades star cluster, tho’ the mainframe model itself is AN SGI ICE X, engineered by SGI.
The mission statement of HECC is “To accelerate and enhance NASA’s mission of house exploration, scientific discovery, and physics analysis by regularly guaranteeing optimum use of the foremost productive high-end computing atmosphere within the world.”
The Pleiades mainframe consists of 161 racks choked with physical science. the essential plan of this kind of mainframe is to string a lot of CPUs along so that they add parallel, all operating along to resolve terribly complicated issues. It runs on a custom UNIX software system.
Within those showing neatness stacked shelves area unit eleven,472 nodes, every containing some variety of CPUs, for a complete of 246,048 central processor cores. That’s a lot of computing power, and it adds up to a peak speed of seven.25 petaflops per second — hundreds of thousands of times quicker than your basic data processor. it’s a complete memory of 938 TB. All this implies that it’s a lot of powerful than we have a tendency to (or most people) will comprehend.
Fun factoid from NASA: “If everybody within the world did one calculation per second for eight hours daily, it’d take regarding a pair of,175 days to complete what Pleiades will calculate in one minute.”
Many people in all probability imagine that you just got to physically travel the mainframe and sit down at a terminal to use it. however that’s not the case.
Instead, the computing resources area unit obtainable remotely, within the cloud. If you enter a science drawback into Google search, it’ll calculate the solution for you. during a similar means, NASA scientists, engineers, and researchers will access the process power of Pleiades from their laboratory or data processor. Time on the Pleiades is precious (which is true of any supercomputer), thus analysisers register prior to to form use of the huge process power to help in their research.
You can really cross-check some cool, easy-to-grasp stats regarding the present usage of the Pleiades, in conjunction with the opposite supercomputers at NASA Ames, at the correct aspect of NASA’s data page. This info is a lot of helpful to researchers UN agency need to use it, however it’s fascinating to visualize a report of this powerful pc in action, updated terribly often (even tho’, if you’re like American state, you don’t grasp precisely what it all means).
Here area unit the CPU’s presently in use within the Pleiades, lessened into the four sets of nodes within the mainframe, that area unit dubbed Sandy Bridge, Ivy Bridge, Haswell, and Broadwell.