Researchers at Stanford's Institute for Computational and Mathematical Engineering (ICME) use HPC as part of a comprehensive research program to better predict flutter and limit cycle oscillations for modern aircraft, understand the impact of turbulent flow on jet engines, and develop computational methodologies that can facilitate the design of naval systems with lower signature levels. ICME needed a cluster to tackle large-scale simulations of flow/structure interaction and aero-acoustic/elasto-acoustic scattering problems.
Join us in discussing how ICME utilized the 164-node Nivation Linux cluster and object-based storage to create a system that is scalable, easy to integrate and manage, and delivers extremely high bandwidth and random I/O.
Rocks Clusters and Object Storage Wednesday, June 29, 2005, 11:00 AM PDT
- Hear the challenges ICME faced in building a cluster capable of handling complex mathematical simulations
- Learn how object storage systems, like those of Panasas, break I/O bottlenecks hindering cluster performance
- Hear Steve Jones' personal experience in managing large-scale Rock clusters and the integration of the Panasas ActiveScale Storage Cluster
- Understand how the joint solution helped accelerate time to results
Highlighted Speakers ============================ Steve Jones, Technology Operations Manager, Institute for Computational and Mathematical Engineering, Stanford University Steve Jones is the sole administrator of multiple Rocks clusters used for sponsored research projects. He has designed and administered a Top500 Supercomputer and speaks regularly about the design and use of HPC systems, most recently as a keynote speaker at the Rocks-a-Palooza Conference at the San Diego Supercomputing Center. His clusters have been featured in eWeek, Bio-IT World, and NPACI's EnVision Magazine. Jones' work has been published in Linux Journal and has co-authored two books related to networking technologies.
Larry Jones, Vice President of Product Marketing, Panasas Mr. Jones has served as a leading force in articulating the product requirements for the Panasas Storage Cluster. Mr. Jones has frequently conducted presentations on issues surrounding cluster computing, parallel file systems and next generation technologies. He will be speaking on the technical requirements of a shared storage solution for a Rocks cluster deployment.
Register Online Today! ================================== There are no fees for this event. To register, just point your browser to:
Sponsored by Panasas =================================== Panasas, Inc. helps companies accelerate the speed and accuracy of their business decisions, enabling customers to maximize the benefits of Linux clusters by breaking down the storage bottleneck created by legacy network storage technologies. The Panasas Storage Cluster combines industry-standard hardware with the Panasas ActiveScale File System and professional services, making Panasas an established leader in object-based, clustered storage. For more information about Panasas, visit