SGI delivers data-intensive solutions:

• USC added a combination of SGI high performance compute and storage solutions for faster time to insight in bioinformatics and medical imaging research.
• Because clusters limited the size and scope of the researchers' experiment models, data searches and genome sequencing efforts, USC purchased the largest shared memory system in the state, an SGI Altix 4700, for the data intensive applications their scientific research required.

"For the data intensive needs of this research we needed a shared memory system, and there aren't many out there that are true shared memory - that was the primary reason we selected SGI Altix", stated Dr. Duncan Buell, chair of the Department of Computer Science and Engineering. "The Altix met our specifications while delivering superior price-performance."

Purchased through SGI-exclusive higher education reseller, James River Technical Inc. (JRTI) with a National Science Foundation grant (PI Dr. Jijun Tang, Co-PIs Drs. John Rose, Homayoun Valafar and Song Wang, key personnel Drs. Edward Gatzke, Thanasis Papathanasiou and Jim Zheng), the SGI Altix system will initially be used for five major research projects, including:

• Defining microbial genome signatures, which measure the evolutionary relationship between species;
• Phylogenetic reconstruction and multiple sequence alignment, a process to determine the evolutionary relationship among organisms and their genomes;
• Protein backbone structure determination using RDC (residual dipolar couplings) data, which focuses on reducing the temporal and financial cost as well as aiming to identify structurally novel proteins;
• Solving computer visualization problems related to the segmentation of 2D and 3D medical image processing;
• Dissecting gene regulatory networks to develop large-scale microarray data analysis incorporating comparative genomics information.

"In addition to the bioinformatics research, another reason we needed a shared memory machine is for medical image processing", added Dr. Buell. "If you're going to track a part of an image that you have identified, from one frame to the next - because people move from one frame to the next and the heartbeat causes registration problems in either video or MRIs - then you need to hold the frames in memory simultaneously. And you need to have space to store high-resolution images, which is why we bought as much storage as we could."

As part of the projects, many researchers will be developing software code on and for the SGI Altix, such as Dr. Jijun Tang and his students. Their phylogenetic research is contributing to building what is known in the genetics community as "the Tree of Life". They will use the SGI Altix to develop and test new algorithms to enhance and extend GRAPPA (Genome Rearrangements Analysis under Parsimony and other Phylogenetic Algorithms), an open-source phylogenetics software programme for reconstructing trees of evolutionary descent.