How Do You Supercharge IBM's Software-Defined Storage Business? Ask Watson


IBM is using its Jeopardy-winning Watson supercomputer technology to push harder into the competitive software-defined storage market space. At an event Monday in Boston, the company is set to debut the first in a series of software-defined storage offerings that will compete with rivals EMC, Google and Amazon.

IBM is pitching its software-defined storage approach as an antidote to the growing enterprise data deluge, in which businesses are now amassing staggering amounts of data on a daily basis. This information historically has been stored by IT in a hierarchical form within a storage architecture, where IT administrators decide where the most critical data will reside so that it can be retrieved more quickly.

IBM believes that's a broken model. And that Watson’s supercomputer technology will help fix the problem.

Vincent Hsu, CTO of Storage Systems, IBM Systems and Technology Group, said the company's new software-defined storage solution, called “Elastic Storage,” is superior to its competitors in that it can shuffle data from hardware to hardware and remote locations using real-time analytics, putting mission-critical data that needs to be accessed quickly on more-expensive flash resources and archived data on less-expensive resources.

"The genius of IBM's solution is as your data patterns change, IBM learns what data is needed where, when and how fast. That eases administration and speeds data access," Hsu said.

Related: Is Software-Defined Storage For Real Or A Lot Of Hot Air?

IBM's new Elastic Storage offering will be the first in a number of software-defined storage solutions rolling out this year, according to IBM. The Armonk, N.Y.-based company said Elastic Storage, coupled with the fact that its new software-defined storage solutions can run on commodity hardware, results in significant cost savings for businesses. Those aggregate savings could reach as much as 90 percent, according to IBM.

IBM's Elastic Storage will be available later this year as a cloud service through IBM's SoftLayer, and will support OpenStack access and application programming interfaces for Hadoop and other big data tools, Hsu said.

"There is big data, and then there is the emerging world of bigger data. IBM's software-defined storage is nothing short of a breakthrough when it comes to the next wave of bigger data quickly overwhelming IT," said Russell Schneider, principal storage consultant at Jeskell Systems, a large government contractor and IBM Premier Business Partner, which has tested the offering.

"IBM can profile the data in real time, expand the storage pools as needed, and intelligently figure out the most efficient hierarchy for data based on real-world usage," Schneider said.

Schneider said storage is a growing imperative for Jeskell's customers, representing about 30 percent of the company's revenue today -- up from 7 percent just a few years ago. "As data continues to drive new business for us in the enterprise, we are about to hit a tipping point where over 50 percent of our revenue will be tied to big data storage needs,” Schneider said. "IBM saw the wave breaking way ahead of EMC, HP and Dell. Now Jeskell is getting ready to ride that big data wave." 

On May 5, EMC unveiled its own Elastic Storage solution called ViPR, designed to unify storage pools spread across hardware and multiple locations. In August, VMware unveiled a software-defined storage solution that ties to the company's VMware vSphere private cloud and virtualization platform. VMware Virtual SAN, or VSAN, helps advance the storage leg of VMware's plan to develop a software-defined data center platform, which will allow customers to build their own Google-like infrastructures.

PUBLISHED MAY 12, 2014