SAS Brings BI Back Home to Big Iron

BI giant SAS announced expanded System z support for its Enterprise Intelligence Platform and touted a new sub-capacity pricing option, to boot

Old mainframe applications don’t die—nor, increasingly, do they fade away, for that matter. Instead, many of them might soon be coming home again, thanks to IBM Corp.’s mainframe revivification efforts.

Case in point: last week business intelligence (BI) stalwart SAS Institute Inc. announced expanded System z support for its Enterprise Intelligence Platform (EIP) and touted a new sub-capacity pricing option, to boot.

SAS was born on the mainframe three decades ago, officials say, and—with the mainframe again vying for enterprise-data-management supremacy—officials claim SAS is the only big BI and analytics vendor to fully support System z.

"SAS has been on the mainframe for 30-plus years, so it only makes sense to ensure that the mainframe can support advances in business intelligence," said SAS senior vice-president Jim Davis, in a statement. "With the SAS Enterprise Intelligence Platform running on IBM System z, enterprises can fully leverage their mainframes to be more agile and strategic with management decisions and feel more confident with their overall information management strategy."

SAS EIP is available for z/OS as well as several flavors of Unix, Windows, Linux, and even OpenVMS. SAS’ EIP comprises not just its BI and analytic/data mining toolset (i.e., its eponymous SAS 9.1 BI Suite), but its data integration stack as well. While IBM and pure-play vendor Informatica Corp. might nab the lion’s share of the ink in the data integration space, SAS’ Enterprise ETL tool is typically ranked among the top three—and has even edged out Ascential, which Big Blue purchased two years ago—for second place in data integration.

SAS’ announcement amounts to a feather of sorts for IBM’s cap. With the introduction of its zSeries Integrated Information Processor (zIIP), Big Blue kicked off an ambitious effort to recast the System z mainframe as a BI- and data management- (DM) friendly platform. With zIIP, IBM was trying out a new—let’s call it "retro"—vision, that of a mainframe-centered Information Utopia.

In this respect, zIIP is a fundamentally different animal than its cousin, Big Blue’s zSeries Application Assist Processor (zAAP). That engine was developed in response to customer demands; with zIIP, IBM is trying to lead customers to an altogether new paradigm. To the extent that ISVs such as SAS give customers additional incentives to buy into IBM’s Information Utopia vision, this also bolsters Big Blue’s case.

The BI Push

This isn’t about subsidizing a moribund mainframe BI and DM market, however, SAS officials insist. Nor is it a question of SAS doing its part to help out long-time partner IBM. Instead, SAS officials claim, there’s an uptick in demand for BI, ETL, and other DM capabilities on the mainframe.

In an interview at February’s TDWI World Conference in Las Vegas, SAS officials didn’t explicitly address zIIP’s contribution to the ongoing mainframe revival—the zIIP engine is simply too new for many prospective customers to wrap their heads around (see http://esj.com/enterprise/article.aspx?EditorialsID=1312)—but they did speak to a general revival of interest in Big Iron application and data access solutions.

After all, many large mainframe shops—particularly federal customers, as well as customers in other public sectors—never moved data off the mainframe in the first place. For this and other reasons, SAS does an increasing amount of Big Iron business, said data integration product marketing manager Ken Hausman. More to the point, Hausman acknowledged, IBM’s mainframe revival—and especially its emphasis on mainframe Linux—has helped spark new interest in the platform among long-time SAS customers.

"We hear from customers all the time. Long-time users [of mainframe SAS]—people that have had SAS [on their mainframes] for years—they’re moving to [Big Iron] Linux or they’re using Linux with their traditional [z/OS] applications, so maybe they need connectivity or maybe they even want to move these [existing SAS] workloads over to Linux," he said.