Mar 27, 2012 (08:03 AM EDT)
Rutgers Gets Big Data Weapon In IBM Supercomputer

Read the Original Article at InformationWeek

12 Hadoop Vendors To Watch In 2012
12 Hadoop Vendors To Watch In 2012
(click image for larger view and for slideshow)
New Jersey's largest university opened a high-performance computing center Tuesday where researchers hope to solve some of the state's biggest big data challenges facing businesses and academia.

At the heart of the new HPC at Rutgers' Discovery Informatics Institute (DII) is a new, $3.3 million IBM Blue Gene/P supercomputer with two racks, a configuration that is capable of providing processing power measured in teraflops. One teraflop is the equivalent of one trillion floating-point operations per second.

Blue Gene/P runs on IBM PowerPC 450 processors in a quad-core architecture. The system has been dubbed Excalibur, in recognition of Rutgers' Scarlet Knights mascot. The center is located at Rutgers' Busch Campus in Piscataway, N.J.

As more funding becomes available, Rutgers plans to add an IBM Blue Gene/Q system, which is the latest generation of the platform. Funding is being provided by university resources, federal grants, and the university's partnerships with IBM and other technology providers.

[ Big data is key to big business. See Will Google Big Query Transform Big Data Analysis? ]

Professor Manish Parashar, director of Rutgers DII, said the system is key to the university's effort to provide researchers and businesses in New Jersey, where some of the world's largest financial and pharmaceutical companies have offices, with access to tools to solve problems through data analysis.

"The hardware provides the communication, storage, and processing capabilities we need to do large data analytics. On top of that, our own researchers have their own software infrastructures that can use this effectively," said Parashar.

Parashar said that some of the applications for the Blue Gene system will be in the areas of modeling flash floods, "which is very relevant in New Jersey," drug design, materials design, and modeling certain cancers so researchers can find better ways to treat them.

"Being able to provide the computational power that can access this data and find the relationship between items, characterize correlations, differentiate patterns--this requires significant memory and compute capacity," said Parashar.

Parashar called the opening of the Rutgers HPC "a huge step in providing this kind of capability. It allows our researchers and students to access not only technology but also to build multidisciplinary research around this and create partnerships with industry."

Parashar said the university and its partners are still determining the business models under which Rutgers' HPC resources will be shared with private industry and other research institutions in New Jersey.

The pay-as-you go nature of the cloud makes ROI calculation seem easy. It’s not. Also in the new, all-digital Cloud Calculations InformationWeek supplement: Why infrastructure-as-a-service is a bad deal. (Free registration required.)