In-Depth

Big Blue Pushes New Grid Bundles, Customer Wins

Morgan Stanley says new computational grid delivers significant new performance in financial analytics.

• By Stephen Swoyer
• 09/30/2003

Grid computing got another shot in the arm last week when IBM Corp. announced a deal with brokerage Morgan Stanley migrate its analytical applications to a grid of Intel-based servers.

The deal, Big Blue’s second with a financial services firm (IBM announced a similar arrangement with brokerage giant Charles Schwab in January), highlights the adoption of grids in non-traditional markets (http://info.101com.com/default.asp?id=154). To this point, grid computing has largely been an academic and theoretical computing play.

Morgan Stanley managing director Richard Anfang says that his company’s new computational grid delivers a significant new performance. "IBM has demonstrated significant performance improvement in a financial analytics application that runs across a Grid of several hundred Intel processor-based machines," Anfang said in a prepared release.

Big Blue touted grid wins with several companies in non-traditional markets, such as human resources outsourcer and consultancy Hewitt Associates. IBM says it worked with Hewitt to build a unique computational grid that supports the latter’s pension modeling application. The grid features IBM eServer zSeries mainframe and BladeCenter technologies working in tandem with GridServer software from DataSynapse, says Pete McCaffrey, program director for zSeries marketing.

“They looked to leverage blade technology for that application and essentially off-load it to a blade environment, and tie this together using grid technology. So by doing that, with this compute intensive app, [they] put it on a technology that’s better suited for it,” he explains. “This enabled them to reduce their transaction costs by approximately 90 percent, while they continue to have an environment that scales. And they managed this whole thing from the mainframe.”

The best part about the grid, McCaffrey claims, is that Hewitt didn’t have to re-write its existing pension application: “They dropped basically a Java call to the grid and the grid dishes the work to available blades, and when it’s done [they] return the results.”

Others adopting grid computing include NLI Research Institute, a company of the Nippon Life Insurance Group, which began a joint research project with IBM's Tokyo Research Laboratory to reduce processing time for the company's financial risk management solution. Big Blue claims that early results have reduced processing time from ten hours to 49 minutes.

New Bundles for Financial Services

IBM last week also unveiled two new grid-based bundles that address the requirements of the financial services industry:

For the first of these, The IBM Grid Offering for Analytics Acceleration: Customer Insight in Banking, IBM tapped software from business intelligence powerhouse SAS Institute Inc. IBM used SAS’ Credit Scoring application as part of SAS’ Banking Intelligence solution to demonstrate that applications can be readily Grid-enabled. SAS was able to exploit the distributed compute power of the grid without customized programming.

The second offering, The IBM Grid Offering for Risk Management and Compliance: Capital Markets and Retail Banking, was developed in conjunction with grid computing purveyor DataSynapse. It’s designed to help risk managers implement a Grid infrastructure to support real-time credit limit monitoring.