IBM uses Watson code for new software-defined storage product

IBM has announced a new storage technology based on software-defined storage. The software is called Elastic Storage and is based in part on technology used by the Watson supercomputer.

With software-defined storage, the software layer is separated from the hardware that is necessary for storing and making data accessible. The software layer should make it easier to run capabilities such as deduplication, replication and backups on large storage systems. With the growth of virtualization and cloud computing, software-defined storage plays an increasingly important role.

Shortly after EMC released its ViPR 2.0 product for software-defined storage, IBM released its Elastic Storage line. According to IBM, its product, which it sells as an iaas service on SoftLayer, is compatible with OpenStack, a popular open source cloud platform, supporting both private and public clouds. Also, Elastic Storage is compatible with the Hadoop API.

According to the IT giant, its software-defined storage product reduces costs for storage systems. The savings could reach 90 percent, for example, because Elastic Storage can automatically move less used data on fast but expensive storage media such as SSDs to cheaper but less fast storage systems, PCWorld writes.

IBM states that the Elastic Storage technology is based on software-defined storage code for its Watson supercomputer. For example, the technology was used when Watson battled two human opponents on the quiz show Jeopardy. In addition, the storage technology managed to copy 5TB of data to Watson’s working memory within minutes. Elastic Storage could address data volumes of thousands of yottabytes of data. One yottabyte is equal to 1 billion petabytes.