Today, during the SAS Global Forum in Orlando, Fla., data warehousing specialist Teradata launched a new Big Data analytics system developed in conjunction with business software maker SAS. Called the Teradata 700 appliance, the RAM-heavy hardware runs pre-installed SAS High-Performance Analytics software to accelerate data modeling for businesses.
According to Tho Nguyen, director of Integrated Solutions for the Teradata-SAS Partnership Program for Teradata, enterprises should notice a dramatic improvement in the time it takes to arrive at business intelligence-backed decisions.
Using a financial company as an example, Nguyen told Enterprise Apps Today that by employing "the traditional way of doing a risk application, it takes 167 hours in the model development process." With the Teradata 700 appliance for SAS High-Performance Analytics, it takes as little 84 seconds, he said.
The secret to such a quick turnaround is RAM, and lots of it.
Taming Big Data with In-Memory Analytics
When it comes to Big Data, in-memory systems excel by storing vast amounts of information in random access memory (RAM), bypassing the performance penalty incurred by fetching data from disk-based storage systems. This allows business intelligence applications to leverage fast server CPUs and system RAM to discover business insights more quickly than is possible using traditional server and storage infrastructures.
Keen on exploiting the profit-generating potential of Big Data, businesses are setting the stage for a healthy market that can improve the fortunes of IT vendors that cater to them. According to a recent IDC forecast, the market for Big Data solutions will grow to $16.9 billion in 2015 from $3.2 billion in 2010.
That means businesses will devote more of their budgets to in-memory analytics in the coming years, an opportunity that's not lost on IT vendors.
Of late, Oracle and SAP have been battling it out to stake out claim in enterprise data centers with their in-memory analytics systems and software offerings. Now with the 700, Teradata joins in with a pre-configured solution of its own that targets SAS environments in data-heavy sectors like telecommunications, retail, finance and banking.
The Teradata 700 system is the result of a "bi-directional partnership" established in 2007 between SAS and Teradata, Nguyen said. It's a partnership that today yields Teradata systems that span up to four cabinets – think server racks -- and contain up to 3,072 GB of RAM.
Available in small, medium and large configurations, the 700 line is built on a software foundation of SUSE Linux 10 and Teradata Database 13.10 in addition to the aforementioned SAS High-Performance Analytics software. Processing power is provided by two, six-core Westmere Xeon CPUs per node operating at 3.06 GHz.
The "small" or base 700 appliance is a two-cabinet model that features 192 processing cores, 1,536 GB of RAM and 20.4 TB of storage. The "medium" model takes up three cabinets and houses 288 cores, 2,304 GB Total RAM and 30.6 TB of storage. Topping off the 700 line is the "large" configuration, with 384 cores, 3,072 GB of RAM and 40.8 TB of storage.