Tech centre in San Francisco will be home to the Spark open-source software project.
IBM has created a new technology centre in San Francisco to work on the Spark open-source software project.
The company has appointed 20 people in the last month, with IBM executive Rob Thomas saying to Bloomberg, "We’re going to be scaling this up to hundreds of people that are just focused on Spark open source and how we evolve that for the enterprise."
Spark is a scalable data analytics platform that features primitives for in-memory computing with some performance advantages when compared to Hadoop’s cluster storage approach.
It offers an efficient framework for the processing of distributed datasets. Spark allows in-memory distributed datasets that optimise iterative workloads apart from interactive queries
Spark, which is implemented in the Scala language, helps firms process vast amounts of information quickly by storing data in the fast memory of computers.
It is considered as an alternative to Hadoop, which attracted several firms including MapR Technologies, Hortonworks, and Cloudera.
Thomas said: "This is a much more significant bet than even what we have done on Hadoop to be frank. We think Spark is going to be enormous and change the face of enterprise IT."
The decreased demand for IBM’s older software and services has forced the company to identify revenue growth from its data analytics business.