SUSE predicts supercomputer capabilities to become part of mainstream IT for enterprise customers

SUSE® says that the SUSE Linux Enterprise Server operating system is being used by more than one-third of the top 100 supercomputers in the world, as ranked by the TOP500 list, a project that tracks and detects trends in high-performance computing (HPC).

SUSE leads in supercomputing through a unique combination of high scalability, flexible architecture and ability to handle large volumes of computer processing units (CPUs). It also validates the overall strength in HPC environments of Linux, which powers more than 95 percent of the TOP500 list.
“SUSE has invested heavily in developing solutions that enable the fastest computers in the world,” said Meike Chabowski, product marketing manager for Enterprise Linux Servers at SUSE. “The next step in supercomputing is to move from high performance to high productivity, and our partners like SGI, Cray and Teradata are helping to lead these advancements. Linux will continue to be an integral part of these revolutionary machines.”

The past few years have seen significant changes in the high-performance computing landscape – often referred to recently as high-productivity computing. This is due at least in part to the emergence of open source and new clustering technologies. The evolution of both lower-cost hardware and Linux has dramatically reduced the cost of these systems. Compute power has increased a thousand times in just a few years, allowing enterprises to use the power of supercomputers in the form of HPC Linux clusters on commodity hardware.

Virtually every industry is adopting Linux clusters to achieve performance improvements needed to deliver on organisational goals. Seismic analysis for oil exploration, aerodynamic simulation for motor and aircraft design, Hollywood special effects, molecular modeling for biomedical research, super-scalable business computing, and data mining and financial modeling for business analysis all leverage HPC. In a silent evolution since 2011, the TOP500 list now includes 270 industrial supercomputers outside academic, research and government use, spanning all industries (with a peak in the Energy/Oil&Gas and IT Service Providers sectors). The fastest industrial supercomputer and number 11 on the TOP500 list is Pangea at Total Exploration Production in France, an SGI® ICE™ X system running on SUSE Linux Enterprise Server.

Bill Mannel, vice president of Servers at SGI, said, “More and more commercial customers are turning to High-Performance Computing to gain the necessary performance they need for their business. With the Pangea system at Total Exploration, SGI developed the fastest supercomputer used in an industrial environment. Partnering with SUSE enables us to use SUSE Linux Enterprise Server, an operating system that uniquely combines high scalability, flexible architecture and the ability to handle large volumes of CPUs.”

Since 1992, SUSE engineers have made significant contributions to the advancement and tuning of the Linux kernel and key kernel-related performance technologies. For example, SUSE Linux Enterprise Server was the first Linux OS in the market to support 64-bit chip sets, and it is synonymous with high-performance Linux running on 64-bit and mainframe systems.

“Since 1998, we have relied on SUSE for our high-performance computer sector at the LRZ,” said Dr. Herbert Huber, head of high-performance systems at the LRZ (Leibniz-Rechenzentrum) of the Bavarian Academy of Sciences in Garching. “At that time, it was important to us that SUSE provided technical features that were not included in other Linux distributions. This is still a huge benefit for us and is why all of our high-performance computers and most other systems run on SUSE Linux Enterprise Server.”

SUSE's early and continued support of new features and technologies relevant to high-performance computing drove and still drives the success of SUSE Linux Enterprise Server in HPC. One area where SUSE invested to support the HPC community and partners is the enablement of Lustre for SUSE Linux Enterprise Server. Lustre is a highly scalable parallel distributed file system which can support multiple compute clusters with tens of thousands of client nodes, tens of petabytes of storage on hundreds of servers, and more than a terabyte per second of aggregate I/O throughput.
Kai Dupke, senior product manager at SUSE, said, “Lustre is a popular choice for businesses with large data centers and high performance needs, including those in industries such as meteorology, simulation, oil and gas, life science, rich media, and finance. By packaging and enabling Lustre for SUSE Linux Enterprise Server, SUSE provides much more value to the HPC community than any other enterprise Linux vendor.”

SUSE engineers are also engaged in storage technologies such as Ceph, a unified, distributed storage system designed for excellent performance, reliability and scalability. Embedded distributed storage platforms like Ceph herald the end of compute and storage CPU silos. The CPUs can be used for what is needed in a certain moment, allowing organisations to maximise CPU utilisation and return on investment.