2007 Research Alliance of Math and Science
Oak Ridge National Laboratory
National Center for Computational Sciences
The National Center for Computational Sciences (NCCS) was founded in 1992 to advance the state of the art in high-performance computing (HPC) by bringing a new generation of parallel computers out of the laboratory and into the hands of the scientists who could most use them. It is a managed activity of the Advanced Scientific Computing Research program of the DOE Office of Science (DOE-SC) and is located at the Oak Ridge National Laboratory.
The NCCS builds on a strong tradition of working closely with computer vendors to bring new high-end machines into productive use and to provide unique, world-class scientific computing resources to DOE-SC researchers in national laboratories, universities, and industry. These resources are vital to DOE-SC’s research programs in basic energy sciences, biological and environmental sciences, fusion energy, and high-energy and nuclear physics. Computing power is key to scientific leadership in these fields, which will provide solutions to many of our most pressing national challenges.
The Research Alliance in Math and Science (RAMS) Program is based on the belief that national laboratories and universities, working hand in hand, offer the best opportunity to make a positive impact on the quality of a diverse workforce. The Research Alliance in Math and Science program is designed to provide collaborative research experiences among faculty and students at colleges or universities and DOE national laboratory researchers. These experiences will improve the U.S. competitive research edge while encouraging and promoting Science, Mathematics, Engineering, and Technology (SMET) research throughout the academic year. In RAMS, I will be working on the jaguar putting in inputs and outputs. Running programs on lustre, optimizing and collecting data from the runs. Jaguar is a Cray XT containing a combination of XT3 and XT4 systems. It is provided as a primary system in the National Leadership Computing Facility (LCF). The current Jaguar installation has a total of 11,706 processor nodes. Of those, 11,508 are configured as compute nodes and the remainder provide I/O and login services. Each of the compute nodes contain 2.6 GHz dual-core AMD Opteron processors and 4 GB of memory. The service nodes consist of a 2.6 GHz dual-core AMD Opteron processor with 8 GB of memory. The system provides an aggregate peak performance of over 119 Teraflops with approximately 46 terabytes of aggregate memory. The system has approximately 600 TB of scratch disk space available for use in the lustre filesystem. Each node is connected to a Cray Seastar router through Hypertransport, and the Seastars are all interconnected in a 3D-torus topology.The resulting interconnect has very high bandwidth, low latency, and extreme scalability. The operating system is UNICOS/lc, which is a combination of Linux on the service nodes and the Catamount microkernel on the compute nodes. Catamount is designed to minimize system overhead, thus allowing scalable low-latency global communication.
Page Designed by: Joylika Adams
