How IBM’s Watson Helped Make Big Data More Affordable

Great publicity is not the only return that IBM realized on Watson, the famed AI that showed us humans just how smart a machine can be.  The IT vendor has incorporated technologies originally developed for the superhuman into entry level servers designed to make analytics much more practical for SMBs.

Big Blue’s new Power Express systems start at $5,947 and come in four configurations: 710, 720, 730 and 740. The servers are powered by the company’s speedy Power7+ chips, featuring 4 to 16 cores depending on the model, and support IBM’s homegrown AIX Unix platform for the Power series, as well as two commercial flavors of Linux.

The one-socket Power 710 Express and the two-socket 730 Express offer up to 5.4 terabytes of capacity in a U2 form factor, with 256GB and 512GB of memory respectively. The 720 and 740 are U4 boxes that feature 7.2 terabytes of storage and a varying number of PCIe slots.

All the Power Express servers ship with core elements of the software that Watson uses to process natural language input. The components include a proprietary analytics engine IBM created for the AI; Hadoop; and additional software offerings, such as Cognos, that are available for a premium.

“When small businesses hear the term Big Data they often think open source and free, but there is still significant cost involved in configuring, deploying and maintaining large-scale clusters of commodity hardware to support Hadoop, for example, both in money and staff time, said Wikibon analyst Jeff Kelly on this morning’s NewsDesk segment with Kristin Feledy (full video below).

“It does require a lot of expertise, and it costs money and time and effort. IBM realized that here’s an opportunity for SMBs that haven’t been able to get involved in big data…to allow them to deploy big data and start using it, at least in a modified fashion, at a price point that is feasible for small businesses,” Kelly says.

Cost is, or was, rather, the big barrier that stood between SMBs and big data analytics. By tearing it down, IBM can better compete with the likes of Dell and HP in the mid market by tapping into the same innovation portfolio that helped it gain an edge in the enterprise.