SGI Builds Out Big Data Capabilities

By Jeffrey Burt  |  Posted 2013-10-22

SGI Builds Out Big Data Capabilities

SGI officials are leveraging the company's high-performance computing heritage to expand their capabilities in the fast-growing area of big data.

The supercomputer maker is offering new solutions that address the escalating compute and storage demands inherent in big data, giving organizations the tools to not only process and store the massive amounts of data being generated and collected, but also to quickly access and analyze it, according to SGI President and CEO Jorge Titinger.

"As businesses tackle the rising volume, velocity, and variety of Big Data, they face a growing challenge - how to unlock value at greater speed, scale and efficiency," Titinger said in an SGI press release. "SGI's expertise in designing and building some of the world's fastest supercomputers enables customers to fully optimize High Performance Computing for Big Data analytics to achieve business breakthroughs."

The new offerings, which include a cluster compute and storage solution, come as more organizations are looking to leverage the vast amounts of data being generated to identify and address potential business opportunities. Big data has become a key trend in business, and a wide range of technology vendors—from hardware makers to software vendors—are looking to meet the growing demand for solutions.

Gartner analysts in March said in a report that adoption of big data technologies are hitting the mainstream this year, with 42 percent of IT professionals in a survey saying they had invested in such technology or were planning to within a year.

"Organizations have increased their understanding of what big data is and how it could transform the business in novel ways. The new key questions have shifted to 'What are the strategies and skills required?' and 'How can we measure and ensure our return on investment?'" Gartner Research Vice President Doug Laney said in Gartner press release in March.

SGI officials said their company's legacy in high-performance computing (HPC) gives it the background to help businesses handle their growing big data needs.

"Today's announcement reflects our rich heritage in HPC and high volume storage, and our ability to significantly help enterprises accelerate time to value, achieve petabyte scale, and lower the rising cost of Big Data," Titinger said.

Among the new offerings is SGI's InfiniteData Cluster, which is powered by Intel's new Xeon E5-2600 v2 processors. It offers up to 12 4TB drives per tray, which creates a 1:1 ratio of processing cores to storage spindles that is optimized for the open-source Apache Hadoop big data software, according to SGI officials. The cluster is coupled with high-speed interconnects, offers up to 40 nodes and 1.9 petabytes of capacity in a single rack and comes integrated with Cloudera's Hadoop distribution running on Red Hat's Linux software and SGI's Management Center tools.

The cluster comes pre-configured and pre-tested, making it faster for organizations to deploy than similar solutions that need to be meshed together once they arrive. It's flexible and scalable, with choices of drives, network connectivity and the number of processing cores, company officials said.

SGI Builds Out Big Data Capabilities

To help enterprises get used to using HPC and Hadoop for big data efforts and to accelerate the adoption of the SGI compute cluster, the vendor has created a Hadoop sandbox that organizations can access online to try out the solution. The sandbox version, which is preconfigured with Red Hat's Enterprise Linux and Cloudera's Hadoop distribution, will be available by the end of the year.

On the storage side, SGI's ObjectStorage offering combines SGI's OEM of Scality's Ring peer-to-peer software and its own Modular InfiniteStorage Server hardware, creating an object-based scale-out storage offering for environments where petabytes of data are being handled. The ObjectStore system creates a shared pool of storage that can support millions of users, offers performance equal to block-based storage and includes the ability to add nodes at any time to increase capacity. The system delivers up to 2.8 petabytes per rack.

SGI officials also will make it easier to index and find infrequently access data that is stored on the company's infiniteStorage Gateway with LiveArc AE, an "appliance edition" of the LiveArc software that will be embedded in the gateway by the end of 2013. LiveArc AE will automatically index metadata and file content, enabling users to quickly search archives via text-free queries to find files, audit trails can keep track of version histories of data and data retention limits can be set, according to SGI officials.

The focus on big data is part of a larger plan by SGI officials to move away from its legacy cloud infrastructure business to focus more time and money on higher-margin efforts, such as HPC, big data and storage. The company announced Oct. 1 that it had bought FileTek, which offers solutions around big data storage virtualization, large-scale data management and archiving.

Titinger said in a conference call with analysts and journalists Oct. 10 that he is pleased with SGI's performance, but admitted that the three-week partial government shutdown could result in the company's revenues for the current fiscal year to fall by 10 percent. SGI has a wide range of government customers, including several military branches and the Department of Homeland Security.

Rocket Fuel