SAS shows off analytics for HPC data jobs

LAS VEGAS — SAS Institute Inc. on Tuesday unveiled new software for high performance analytics which will work on EMC Corp.’s Greenplum Data Computing Appliance family.

Showcasing the new product at this week’s SAS Global Forum 2011 in Las Vegas, the Cary, N.C.-based business intelligence giant said that bringing high performance analytics to a data warehouse appliance could allow analysts to run massive analytics problems within minutes or seconds as opposed to weeks or days. The product will be geared toward verticals such as health care, retail, government, pharmaceuticals and financial services.

Keith Collins, senior vice-president and chief technology officer with SAS, said the software is a “game changer” for in-memory processing. Moving beyond in-memory databases, the software can handle complex analytics jobs by breaking up the task and running it in a parallel, shared memory environment, he added.

Processing a billion rows of data, which could haven taken up to 20 hours, can now be completed as quickly as eight seconds, said Collins, adding that analytics should never be limited to the type or size of data a company is analyzing.

The new software is part of the company’s larger high computing portfolio, which includes in-database analytics, grid computing and in-memory analytics. It will be available on EMC Greenplum appliances in the fourth quarter.

SAS also announced a partnership to bring its high performance analytics software to the Teradata appliance.

One company which has shown a major interest in high performance analytics is U.S.-based retailer Macy’s Inc. The retailer, which showcased its partnership with SAS at the conference’s opening session, has worked with SAS to build a high performance markdown optimization solution.

Macy’s is using the analytics tools to crunch data from more than two million clothing items across 800 store locations.

Determining how to best markdown seasonal clothing items across various regions is an impossible task without heavy duty analytics, said Brian Leinbach, senior vice-president of systems development and applications at Macy’s. He added that the company is even culling weather data in different regions of the U.S. to determine how to best price its knit products in the winter.

Other large retailers were also in attendance at the show. Jody Crozier, divisional vice-president for merchandise solutions at Matthews, N.C.-based Family Dollar Stores Inc., said that until recently his organization has been run like a national chain with a “one-size fits all” approach across its 7,000 store locations.

With analytics technology from SAS, Crozier said the company has changed its approach and started laying the foundation to do localized assortments of its product displays and store layouts. The retailer is analyzing data to determine the various needs of customers based on factors such as urban versus rural, income, and the ethnic makeup of the surrounding area.

“With 7,000 stores, this couldn’t be executed manually in one spreadsheet,” he said, adding that the next big analytical project to tackle will be price optimization.

Chris Vukich, vice-president of pricing at Jacksonville, Fla.-based grocery retailer Winn-Dixie Stores Inc., said one of the most difficult challenges for any technology or business leader tasked with high performance analytics projects will be to gain corporate support.

“For associates with 30 or 40 years of tenure, when you ask them if they use analytics, they’ll say ‘Sure, I use Excel,’” he said.

At Winn-Dixie, Vukich opted against undertaking a one-off pricing optimization project and decided to package together a comprehensive “science of retail” plan, which showed how analytics could help revamp the entire product pricing scheme.

“When we presented the whole thing together, we were able to get senior executives to buy-in,” he added.

One competitive advantage that analytics has brought to the retailer is the ability to more effectively classify product demand across the country, right down to the city level. For example, the company has discovered plantains should be grouped as a high demand product in Miami, but a lower demand product in nearby Jacksonville.

Related Download
3 reasons why Hyperconverged is the cost-efficient, simplified infrastructure for the modern data center Sponsor: Lenovo
3 reasons why Hyperconverged is the cost-efficient, simplified infrastructure for the modern data center
Find out how Hyperconverged systems can help you meet the challenges of the modern IT department. Click here to find out more.
Register Now