In recent years, high performance computing has had a major impact on the world and the development of IT systems in general. The nature of weather forecasting has changed to be much more precise. Our lives have been lengthened by the types of research that medical scientists are able to perform, and we now can see through clouds and space dust to determine what black holes are made of.
This week at the Supercomputing Conference (SC’14) in New Orleans HPC will take center stage. As technology evolves, the role HPC plays in solving the world’s toughest problems becomes reality every day.
Seagate is proud to be one of the companies that has been supporting the evolution of technology for more than 30 years. As Ken Claffey, vice president and general manager of the company’s Cloud Systems & Solutions division, recently commented, “At Seagate we believe in working closely with the pioneers that will help shape the future of IT. Our belief is that the HPC community is really at the forefront of defining the future.”
Of all the elements that come into play for successful HPC implementations – no matter what the scale – storage is one of the most critical. Storage technology can either be a bottleneck or an enabler. Seagate has delivered a string of industry firsts that have helped define state-of-the-art scale-out architecture for HPC storage systems.
Seagate’s innovation is extended through the device to the information infrastructure. This brings a unique capability unmatched by any of its competitors – continuum of ownership across the component chain. The company designs and manufactures its components and drives, firmware, storage enclosure, software enclosure management and software layer. This is demonstrated best in ClusterStor, a fully integrated scale-out solution designed for HPC, cloud and big data customers. ClusterStor delivers what Seagate likes to call the “three Ps” – performance, productivity and profitability.
- Performance – ClusterStor’s architecture provides unmatched speed and efficiency. The high-end ClusterStor 9000 and its family members, the ClusterStor 6000 and ClusterStor 1500, are also the most efficient HPC storage systems on the market. Their architecture allows the systems to achieve greater performance per disk drive, making the entire rack faster and more efficient.
An example of performance is what our strategic partner, Cray, is doing with Seagate storage technology in the contract awarded by the National Nuclear Security Administration. The $174 million contract includes Cray’s recently announced XC40 supercomputer and the new Cray Sonexion 2000 storage system line. The Cray Sonexion 2000 is a scale-out Lustre storage system, purpose-built for big data and supercomputing, Powered by Seagate. The parallel file systems used for primary storage will be comprised of numerous racks of Cray Sonexion 2000 storage. The integrated solution will include approximately 82 petabytes of total usable capacity and about 1.7 terabytes-per-second of total aggregate performance.
- Productivity – Seagate ClusterStor products are designed for ease of use and management. Before being shipped, each system is fully integrated and tested at the factory. This means that users can be up and running in hours vs. days, weeks, or even months required by competitive offerings or do-it-yourself “science project” options. A rapid installation, combined with ease of use means faster ROI as users speed up their time to results for their HPC science, big data or other projects.
- Profitability – Rapid ROI is part of a larger picture – users are more profitable due to the efficient nature of ClusterStor’s architecture. Seagate is able to optimize ClusterStor at every layer of the storage stack – from the application to the file system, all the way down to the storage media itself. This integrated approach means fewer disks, less supporting infrastructure, power and cooling savings that can run as high as 33%, and reduced administrative and overhead costs. Not only is ROI positively impacted – total cost of ownership is reduced as well. When dealing with HPC systems, the savings can be dramatic – sometimes hundreds of thousands of dollars in operating costs alone.
Seagate believes in working closely with the pioneers in HPC who will shape the future of high performance computing and ultimately every aspect of information technology. Given its leadership position and commitment, it’s no wonder the company’s motto is “Seagate is HPC storage!”
Stop by our booth (#3239) at SC and/or follow Seagate Cloud Systems and Solutions on Twitter @SeagateB4B and LinkedIn