ACADEMIA
Panasas Extends Dominance In University Cluster Installations
Panasas, Inc., today announced that Louisiana State University has joined an ever-growing list of the nation's top institutions of higher education leveraging the Panasas ActiveScale Storage Cluster and high-performance Linux clusters to conduct important scientific research. The Panasas Storage Cluster has also been deployed at leading universities such as Stanford, Princeton, Carnegie-Mellon, Rockefeller and multiple campuses from the University of California system to support world-class scientific, bioscience, and computational research. The scientific efforts at these universities help gain a better understanding of the world around us, bring life-saving medicines to market faster, and improve the overall quality of peoples' lives. "Universities are deploying high-performance Linux clusters for research that has a profound and demonstrable impact on the world around us. Until now, complimentary storage solutions have created a performance bottleneck that limited cluster performance and ultimately slowed the research efforts," said Victor Perez, president and CEO of Panasas. "We understand the computing needs of the academic community and we welcome the opportunity to show other leading universities how our innovative storage solution can address their very stringent requirements." LSU's Center for Computation and Technology (CCT), which is housed at the Louisiana State University's Baton Rouge campus, has deployed 32 terabytes (TBs) of the Panasas Storage Cluster. By leveraging the Panasas ActiveScale File System and DirectFLOW client to deliver direct disk to cluster node access, researchers are able to manipulate very large data sets with an 8X performance increase over existing NFS servers. The Panasas system has also met LSU's scalability and management requirements, enabling LSU to deploy more capacity as network usage requires with no downtime for the users or degradation of cluster performance. "In terms of price/performance and total value, Panasas is the best storage solution we've found to meet the evolving needs of our HPC environment," said Brian D. Ropers-Huilman, assistant director, HPC and Computation Center for Computation & Technology at LSU. "With the Panasas solution we have significantly increased the speed with which we conduct our research projects, and maximized the overall productivity of our researchers." Accelerated Time To Results For University Research By combining a parallel file system with object-based storage, the Panasas Storage Cluster delivers scalable capacity and bandwidth in a single, seamless namespace. The integrated hardware/software platform is easy to deploy and enables customers to break the bottleneck inherent in legacy network storage systems. Additional universities leveraging the Panasas Storage Cluster include: Stanford University Institute for Computational and Mathematical Engineering The Institute is using Panasas storage in conjunction with NPACI Rocks cluster software and a Linux cluster as part of a research program to better understand the impact of turbulent airflow (flutter) on turbine engines. Such research is aimed at improving the performance and reliability of jet engines, improving air quality from engine emissions, and improving the noise quality for communities near airports. The Panasas Storage Cluster dramatically increased system performance by eliminating the I/O bottleneck caused by an existing NFS storage system. Researchers now take only an hour to complete jobs that used to take an entire day to process. University of California-Berkeley Center for Integrative Genomics At the UC-Berkeley Center, researchers from traditionally isolated fields of study use the Panasas Storage Cluster in conjunction with a Linux cluster to enhance comparative genomic analysis to understand the complex mechanisms that create evolutionary diversity among animals, plants and micro-organisms. Using the Panasas Storage Cluster, research teams have been able to employ breakthrough approaches to parallel data queuing and distribution. University of California-Santa Cruz Center for Biomolecular Sciences and Engineering This Center leverages the KiloKluster, a 1000-node second-generation bioinformatics cluster that processes large quantities of biological and biomedical data that have arisen from genomic sequencing. The Panasas Storage Cluster was selected over competitive storage products that failed to meet the very high scalability, performance and reliability demands of the KiloKluster system. University of California-San Francisco Andrej Sali Laboratory Using the Panasas Storage Cluster, researchers at the Sali Laboratory have realized a 5X improvement in performance of key applications that study the structure and function of proteins. Such research can lead to accelerated development of life-saving medicines by understanding how diseases form at the molecular level. Panasas was chosen over other storage solutions for its random I/O performance and ease of management through a single global namespace. Rockefeller University, Gaasterland Laboratory of Computational Genomics The Gaasterland Laboratory of Computational Genomics aims to create and use new software tools to explore the unfolding world of genomes. The Panasas Storage Cluster was selected for its ability to scale seamlessly without disruption to current users, provide fast data throughput by eliminating the I/O bottleneck traditionally experienced with legacy network storage solutions, and deliver extreme reliability and availability to an extended community outside the Lab. Organizations interested in learning more about Panasas' programs for universities can visit its Web site.