By Madan Sheina

Netezza is preparing to embed the compression engine technology it unveiled last December into expanded configurations of its core data warehouse appliance system in order to ratchet up data scalability and query performance.

Netezza said its Compress Engine will be included as part of its entire NPS (Netezza Performance Server) 10000 series range of appliances later this year to bump up data scalability from tens of terabytes up to hundreds of terabytes and even petabyte levels. The new configurations will also speed up data loads and backups to let the NPS appliances fly more easily in 100 terabyte-plus environments.

Phil Francisco, director of product marketing at Framingham, Massachusetts-based Netezza, said the move to scale-up its NPS appliances for higher capacity performance reflects a growing number of customers that are now looking to build and run analytic applications against even larger data sets.

Francisco regards data volume growth rates of 30-100% per year as being commonplace in industries like financial services, telecommunications, online businesses and federal government.

"All of these are key verticals for high-end data warehousing and for Netezza. They're all looking to perform advanced analytics against larger quantities of data which is pushing warehouse sizes up," he said.

Francisco added that it was important for Netezza to make its customer base feel "comfortable" about their appliances' ability to step up to plate.

"We're giving these customers assurance that our NPS appliances will be able to scale from several gigabytes up to multiple terabytes.

The Compress Engine that Netezza took the wraps off last month, along with additional technologies that increase the data density of NPS appliances, are key enabling technologies to make that happen, Francisco said. "It's a big performance multiplier for us."

The Compress Engine basically introduces compression algorithms that cleverly and efficiently compile column-based data tables to disk to maximize access and query performance. Netezza claims the engine can double query processing performance overnight.

The engine is the latest addition to Netezza's new FAST Engines framework which uses field programmable gate arrays (FPGAs) and other commodity components of its appliance to push processing power as close to the data as possible. It joins four other FAST "engines" –- for managing access to data flows, breaking down data streams into relevant rows and columns, ensuring data integrity and efficiently filter row and column data into memory or to specific queries.

The FAST Engines continue a theme of scalability and performance improvements by Netezza, Francisco said.

He explained how over past several years Netezza has continually pushed the envelope in terms of top-end performance and scalability innovation. "In 2002 the company's NPS platform scaled to 6 terabytes across a 4-rack system. In 2004 Netezza bumped that up to 27 terabytes in a six-rack configuration and further extended that to 100 terabytes in 8 racks in 2005."

"The planned expansion of our NPS product line this year is really a continuation of our strategy of extending data density and capacity and the range of analytic applications we can support. It's really the next logical step for us," Francisco said.

Francisco expects the new compression-enabled NPS configurations to be available sometime in May 2008.

Our View

Ever since Netezza decided last summer to recast its NPS appliance from a data warehousing platform to a fully-fledged analytic system for both specialized and general high-end analytics, it has focused on improving data density and performance.

The net result of that was the introduction of its so-called FAST Engines, which provided customers with a way to incrementally supercharge their data warehouses for high-end analytics against growing mountains of business data. The engines provide the processing muscle to allow Netezza to stick to its core design philosophy of moving analytic processing as close to where the data is physically stored, even in the face of escalating data volumes. The compression technology that Netezza has developed, however, seems to be different from traditional techniques used by relational database providers –- the main difference being that it also addresses performance and not simply reducing data size.

With a successful IPO behind it Netezza is brimming with confidence and clearly has its eye on much bigger analytic problems that when it first started out. The early messaging of its NPS appliances being a complementary adjunct to large Oracle, IBM and Teradata enterprise data warehouses for specialized complex analytics is starting to wear thin, especially as it now boasts a cheaper and easier to deploy systems that now also scales up to the same analytic challenges. Make no mistake the FAST Engine-enable NPS appliances are now taking aim at high-end enterprise data warehousing accounts.