SAS Makes Triple PlaySAS Makes Triple Play

5 Big Wishes For Big Data Deployments

SAS unleashed three major announcements at its Global Forum in San Francisco this week, with a wave of new SAS High-Performance Analytics capabilities, a unified SAS Customer Experience marketing suite and added levels of support for public and private cloud deployment of SAS software.

SAS High-Performance Analytics software is designed to take advantage of highly distributed, massively parallel processing (MPP) on memory-intensive X86 servers. It has been a big strategic push for SAS over the last two years as customers demand ever-faster performance.

"The windows of opportunity to compute analytics are getting shorter and shorter," SAS CEO Jim Goodnight recently told information in a preview of the Global Forum announcements. "People now want calculations almost in real time with almost everything that we do."

These low-latency demands cut across financial services, retailing, life sciences, oil and gas exploration, manufacturing and other industries. And they're often accompanied by ever-growing volumes of data from new risk, customer behavior, genomic, remote sensor and supply chain measures.

[ Want more on recent SAS data-visualization moves? Read SAS Upgrades Visual Analytics App. ]

SAS previously offered High-Performance versions of industry specific applications such as financial risk analysis and marketing optimization. But with the SAS High-Performance Analytics upgrades announced this week and set for release in June, SAS will bring MPP power to six core products in its portfolio: Statistics, Data Mining, Text Mining, Optimization, Econometrics and Forecasting.

These general-purpose capabilities are applicable to most any industry and were heretofore sold together with standard SAS software. But as part of SAS High-Performance Analytics, the modules can be purchased individually to address the most demanding applications. In this way, customers can step up to high-scale or processing-intensive analysis in an affordable fashion, according to SAS CTO Keith Collins.

"It lowers the entry point and lets people focus on the techniques where there are bottlenecks at scale," Collins told information. "That might mean Modeling for customer rewards analysis, Optimization for inventory optimization or Forecasting for retail, as an example."

SAS makes the case that it can address big data variety as well as sheer volume because SAS High-Performance Analytics can now run on Hadoop clusters. Customers need only use SAS' Anyfile Reader utility to map data on Hadoop cluster to SAS.

"We're able to map to and manage the metadata around any file in Hadoop, whether it's JSON, XML, comma-delimited or binary," Collins explained. "If you dump log files into Hadoop, we can also map to those [semi-structured] formats and use them with traditional analytics." The latest release also brings Oracle Database into the fold as a partner platform on High-Performance Analytics, with many in-database processing options along the lines of those previously supported only by EMC Greenplum and Teradata.

SAS speed-of-analysis options now range from in-database to in-memory to in-stream, with the last option supported by SAS DataFlux Event Stream Processing technology. "Some things you'll want to do inside the database, like scoring," Collins explained. "You'll want to handle some processing-intensive analytic computing in memory, and the real-time work you'll want to handle with streaming technology."

In an example of a streaming application, Collins said a large bank has created a consolidated view of myriad risk-analysis computations. Stream-processing technology is used to connect, run and continuously monitor the risk measures.

In the area of marketing, SAS has been successful with purpose-built applications including SAS Marketing Automation, SAS Marketing Optimization, SAS Real-Time Decision Manager and SAS Digital Marketing. But marketers have been asking SAS for a more cohesive combination of these capabilities as they seek to develop and optimize cross-channel marketing campaigns.

With this week's release of a redesigned SAS Customer Intelligence product, the vendor is exposing all four applications within a single, integrated product. And it wasn't a simple matter of putting a portal-like veneer with single-sign-on capabilities in front of four otherwise-unchanged products.

"We spent a tremendous amount of time and energy working with customers to understand their processes, so it's a complete rethink about how work flows," Collins said.

Customers can chose any or all of the four marketing modules that are relevant to their needs, and the functionality shows up within process-oriented workflows, according to Collins. Rather than moving from application to application, users move more naturally through sequential and related tasks.

Choice is the theme behind the SAS cloud announcements this week, and the options include "your cloud, my cloud or their cloud," Collins said, meaning customer private clouds on virtualization technology, SAS managed services or third-party public clouds such as Amazon Web Services.

These choices are facilitated by a SAS 9.4 foundation software release set for June that incorporates cloud-friendly architectural enhancements including automation features developed to facilitate virtualized capacity deployment. The technology was gained through the November acquisition of cloud infrastructure company RPath.

Big data, technology-driven marketing and cloud deployment are, of course, trendy topics for lots of tech vendors. But SAS, as always, has focused on advanced analytics and what these enablers can do to deliver its tag line "power to know" that much more quickly, flexibly and in-line with day-to-day business processes.