ACADEMIA
NASA Scales Pleiades InfiniBand Cluster to 25,000 Cores
- Written by: Webmaster
- Category: ACADEMIA
SGI announced that with over 60 miles of InfiniBand cabling in place at the NASA Advanced Supercomputing (NAS) Division at NASA Ames Research Center at Moffett Field, Calif., a scientist was able to utilize 25,000 SGI ICE Intel Xeon processor cores on Pleiades to run a space weather simulation.
One particular area of study is magnetic reconnection, a physical process in highly conducting plasmas such as those that occur in the Earth's magnetosphere, in which the magnetic topology is rearranged and magnetic energy converted to kinetic or thermal energy. This field of research is critical, as these disturbances can disable wide scale power grids, affect satellite transmissions and disrupt airline communications.
“We study the efficiency of reconnection under different conditions in the magnetosphere,” said Dr. Homa Karimabadi, space physics group leader at the University of California, San Diego. “This includes such things as the amount of mixing of the plasma from the solar wind and from the magnetosphere. The team then uses the local-scale details to improve models of magnetic reconnection in ‘global’ simulations encompassing a region starting at about three earth radii andextending to 30 to 200 earth radii, with the Earth’s radius being about 6400 kilometers.”
As detailed in the article, “Cracking the Mysteries of Space Weather,” by Jarrett Cohen of NASA, Earth is mostly protected from solar flares, coronal mass ejections, and other space weather events by the magnetosphere, a magnetic field cocoon that surrounds it. But sometimes Earth's magnetosphere ‘cracks’ and lets space weather inside, where it can cause damage. Getting space weather details right means capturing everything from the 1.28 million kilometer-sized magnetosphere down to subatomic-scale electrons. Doing that in one simulation would require supercomputers more than 1,000 times faster than those available today, so the research team breaks the problem into two parts. They start with‘local’ simulations that include full electron physics of regions in the magnetosphere where reconnection is known to occur, followed by ‘global’ simulations.
Accessing up to 25,000 processor cores on Pleiades, Dr. Karimabadi said that his group can run ‘kinetic’ simulations that treat each electron with its full propertiesand understand how electrons allow reconnection to occur. In the local simulations, electrons are treated as individual particles. In the global simulations, electrons are treated as fluids and ions (electrically charged atoms) as particles.With Pleiades, simulations can run for five days straight, enabling many parameter studies. Among recent findings is that magnetic reconnection by itself is quite turbulent, producing vortices in the plasma that create many interacting flux ropes—twisted bundles of magnetic field. As observed by spacecraft, flux ropes can extend several times the radius of Earth.
Space weather is just one application running in the petascale range today that will benefit from the move to Exascale computing. SGI is committed to bringing Exascale solutions to the marketplace in an open computing paradigm in which InfiniBand will have an integral role by providing the key interconnect elements. Having demonstrated that InfiniBand can enable petaflop-sized systems at NASA, SGI will also be partnering with Total to bring InfiniBand into the multi-petaflop commercial space. Coupled with software tools such as SGI Management Centerand SGI Performance Suite, as well as big data InfiniteStorage solutions, SGI is positioned to offer an optimal user experience for multi-petaflop deployments moving to the exascale range.
“The growing complexity of problems that Pleiades is expected to solve is why we remain committed to developing and providing exascale solutions,” said Praveen K. Mandal, senior vice president of engineering at SGI. “NASA has been a long-time partner of SGI in pushing the boundaries of technology development andscientific discovery. For five generations, SGI ICE has tightly integrated InfiniBandtechnology to achieve top performance and scaling capabilities in a completely open networking architecture, and at the same time, with five different networking topologies, ICE is flexible enough to fit all the workloads that benefit science.”
Maximizing productivity in today’s HPC cluster platforms requires using enhanced data messaging techniques. On Pleiades, every single node has a “direct” InfiniBand connection to the rest of the network. That is why Pleiades has the largest InfiniBand network of any HPC system in the current Top 500 list. Byproviding low-latency, high-bandwidth, and a large message rate, high efficiency interconnect solutions are used as high-speed interconnects for large-scale simulations such as these conducted by NASA, and are replacing proprietary or low-performance solutions.
“The scalable architectural design of the SGI ICE platform provides an excellent vehicle for showcasing the performance, scalability and efficiency of Mellanox InfiniBand server and storage interconnect solutions,” said David Barzilai, vice president of marketing at Mellanox Technologies. “Utilizing InfiniBand’s ability to efficiently scale to 10,000 and more server nodes, Pleiades is able to run space weather simulations of greater complexity faster and with more precision.”