Japanese Genomics Organization Chooses New SGI HPC System
By Allison Proffitt
November 1, 2011 | The Institute for Chemical Research (ICR) at Kyoto University in Japan, has selected an SGI UV 1000 high performance computing and storage system to advance its genomics research and enable GenomeNet, a network of database and computational services for genome research and related research areas in biomedical sciences, operated by the Kyoto University Bioinformatics Center.
The solution will consist of SGI UV 1000 systems operating at 32.6 teraflops and delivering up to 6.6x performance increase over the previous SGI system that was in place. The new solution was configured with over 3072 cores of Intel Xeon processor E7 series. It will contain 48TB of memory and include 840TB of storage.
The SGI UV 1000 system, previously marketed as Altix UV, “can really tackle the data intensive challenges [ICR] has in their life sciences and chemistry work,” Jill Matzke, director of server product marketing for SGI, told Bio-IT World. The system has “very large caches of memory so very large datasets can be accessed very quickly.”
“SGI UV has been chosen by a large number of life sciences researchers in the genomics domain,” Matzke explains, “because it’s a really accessible and easy to use system. Genome assembly runs really well and quickly on this type of platform. Because this institute has one of the more demanding research databases, they’ve purchased a rather large system.”
Matzke says the system is adept at handling not just sequencing, “which everyone can do these days,” but also the computational challenges of applying sequence data to characterize biological systems.
“Our research laboratory developed the Kyoto Encyclopedia of Genes and Genomes (KEGG), an integrated biosystems database which has been providing services to researchers around the world as part of GenomeNet,” said Minoru Kanehisa, professor and director of the Bioinformatics Center, in a press release. “Adopting the SGI UV 1000 for GenomeNet calculations will dramatically reduce processing time to provide users with the latest data as quickly as possible, and we look forward to taking full advantage of the large amount of genomic information generated by next-generation sequencers as a center for shared usage and joint research.”
Twin Servers
The system consists of two servers: one for computational chemistry and one for GenomeNet calculations. The computational chemistry server consists of two nodes with 512 cores and 8TB of shared memory. Applications such as quantum chemistry and molecular dynamics will be utilized on the UV 1000, enabling users to run Gaussian, CASTEP, Discover and other programs to examine molecular structures and their specifications and characteristics. The GenomeNet calculation server also consists of two nodes with 1024 cores and 16TB of shared memory.
Matzke says that SGI is very focused on life sciences thanks to recent “explosive growth” and increased investments. Genomics is an obvious area of potential growth, but Matzke says medical image research is another area generating huge databases. She says that SGI is a good fit for organizations like ICR that have “really top of the heap kinds of problems that run of the mill solutions can’t handle.”
The ICR product is one of the larger deployments of a standard product, said Matzke, so the configuration has been scaled up for their deployment. It’s a hands-on approach that Matzke believes serves SGI well. “Customers like Kyoto ICR have access to specialized capability and they can actually get to the resources because we’re not a huge 'megatron' organization,” she says.
The new system should be in place by the end of 2011.