ACADEMIA

Application period for SuperCompute time is September 15 through October 15, 2012

The San Diego Supercomputer Center (SDSC) at the University of California, San Diego is seeking innovative applications for the next round of user allocations on its Gordon supercomputer, which went into operation earlier this year.

SuperCompute time on Gordon, the first supercomputer to use massive amounts of flash-based SSD (solid state drive) storage, is available through the National Science Foundation’s (NSF) Extreme Science and Engineering Discovery Environment, or XSEDE program. More information on the allocation process can be found at xsede allocations

Applications for time on Gordon will be accepted during the one-month period starting September 15 and ending October 15. Awards will be announced following the December meeting of the XSEDE Resource Allocations Committee (XRAC), which consists of scientists who review the submissions and make allocation recommendations for all XSEDE resources. SuperCompute time for this round of allocations is awarded for one year commencing January 1, 2013.

Gordon contains 300 TB (terabytes) of flash-based storage. Gordon also uses large memory supernodes that aggregate multiple x86 nodes into a single virtual system, delivering a high-end symmetric multiprocessor (SMP) supercomputer.

The standard supernode has approximately 1 TB of DRAM, but larger memory configurations can be arranged as needed. These features make Gordon an ideal platform for tackling data-intensive problems. SDSC’s Gordon is the result of a five-year, $20 million NSF Grant.

Gordon is a new Appro Xtreme-X supercomputer that helps researchers tackle the most difficult data-intensive challenges, from mapping genomes for personalized medicine to rapidly calculating thousands of “what-if” scenarios affecting everything from traffic patterns to climate change.

Gordon is a dedicated cluster consisting of 1024 compute nodes and 64 I/O nodes. Each compute node contains two 8-core 2.6 GHz Intel EM64T Xeon E5 (Sandy Bridge) processors. The I/O nodes each contain two 6-core 2.67 GHz Intel X5650 (Westmere) processors, 48 GB of DDR3-1333 memory, and sixteen 300 GB Intel 710 solid state drives. The network topology is a 4x4x4 3D torus with adjacent switches connected by three Mellanox 4x QDR InfiniBand links (120 Gbit/s). Compute nodes (16 per switch) and I/O nodes (1 per switch) are connected to the switches by Mellanox 4x QDR (40 Gbit/s). The peak performance of Gordon is 341 TFlop/s.

“Gordon is already assisting the research community in a wide range of data-intensive projects, many of which could not be addressed previously because scientists simply didn’t have the computer capability to do so,” said SDSC Director Michael Norman, principal investigator for the Gordon project. “We are looking to attract researchers who have innovative larger-scale research projects that will utilize Gordon’s unique capabilities.”

The system’s supernodes are ideal for users with serial or threaded applications that require significantly more memory than is available on a single node of most other systems, while Gordon’s flash-based I/O nodes may offer significant performance improvement for applications that exhibit random access data patterns or require fast access to significant amounts of scratch space.

In addition to the standard XSEDE allocation, SDSC is also inviting researchers to apply for dedicated use of Gordon’s flash I/O nodes for up to one year. This provides them with a platform for new types of data intensive applications, for example hosting community data resources that are closely integrated with compute nodes, or complex workflows where instrument or simulation data is staged on the flash nodes, and subsequently accessed for analysis and visualization. Dedicated I/O node allocations are processed through the XSEDE Start-Up allocation request, which can be turned around in a matter of a couple of weeks.