High Performance Computing—Powering Mars Missions and Handling Big Data
by Paul Rudo on 01/10/12 at 9:23 pm
NASA’s Mars Curiosity rover has been on Mars for two months now, and in that time has managed to cover all of about 800 feet of terrain (this figure becomes much more impressive when you consider that Curiosity’s younger cousin, Opportunity, has been on our neighbor planet for eight and a half years and has only traveled 21.78 miles from its landing site). Along the way, it’s been sending back some spectacular photographs from the surface of the Red Planet and is currently searching for evidence of environments that may once have played host to microbial life.
Curiosity is also paving the way for future manned missions to Mars by collecting more data about conditions on the surface of the planet. As such, Curiosity plays a critical role in helping to determine NASA’s upcoming focus and objectives.
Curiosity Supported by High Performance Computing
Curiosity’s mission would probably not have been possible without massive computing power to back it up. Two months ago, when the world’s eyes were riveted on the rover’s initial touchdown on Mars, much was made about the so-called “Seven Minutes of Terror”—the period during which Curiosity was entering the planet’s atmosphere and would lose contact with NASA. The landing sequence was complex—it took place in three stages and, among other things, involved the largest supersonic parachute ever constructed and a sky crane.
Needless to say, a successful landing was, by no means, guaranteed. There were hundreds of variables that had to be accounted for and thousands of calculations and tests that had to be carried out before the landing could take place. And that’s where high performance computing (HPC) comes in.
Essentially, high performance computing is a supercomputing concept that utilizes parallel processing to run advanced applications that would otherwise be too large or too complex to compute. In other words, HPC “clusters” link together a network of computer nodes that each contains one or more processor and its own memory. These parallel nodes work together to solve tasks that would simply take too long for a single computer to process.
For its landing sequence (certainly the most complicated part of the mission), NASA’s Jet Propulsion Laboratory relied on two Dell high performance computing clusters (unsurprisingly named Galaxy and Nebula) to conduct analysis on the large amount of test data necessary to properly prepare Curiosity for the Seven Minutes of Terror. In fact, the final landing sequence parameters were uploaded to Curiosity only a week before its big day.
High Performance Computing Isn’t Only for Aerospace Engineers
The applications for high performance computing go far beyond missions to Mars, however. The same HPC clusters that are used to calculate landing trajectories and factor in Red Planet weather patterns can be used by businesses and other organizations to analyze and store their big data.
With the mountain of data now being collected by many companies, standard servers just do not have the necessary processing power to handle the load. The powerful processing power offered by HPC allows businesses to sift through tera- and petabytes’ worth of information to gain valuable marketing, sales, and trends insights. Though they may not be computing anything as exciting as a Mars rover landing sequence, organizations that utilize HPC technology are able to remain at the forefront of innovation.
Many organizations are harnessing HPC power to propel new discoveries and results, speed time to market, and gain a better understanding of their customers’ needs.
HPC Accessibility
One of the primary benefits that high performance computing has over traditional supercomputing is that it is completely scalable and available even to smaller firms. Since HPC clusters are made up of many nodes working together, computing power can be added or removed as necessary. This allows companies to design computing solutions that can accommodate their growth and handle their specific requirements.
In short, high performance computing makes some pretty amazing things possible. Whether it’s piloting a small land rover millions of miles through the vacuum of space or filtering through sales data to determine what customers may be lacking, HPC in helping people solve problems and think about the world in new ways.
Author Bio: David Malmborg works with Dell. When he isn’t working he enjoys hiking, spending time with family and researching new technology. Learn more about high performance computing here.
Recent Comments