British in-memory analytical database firm Kognitio announced the launch of version 1.0 of Pablo, saying it enables ordinary users to perform "train of thought" analysis by building and querying virtual OLAP cubes on the fly, and giving Excel users the ability to query Terabytes of data.
Kognitio’s flagship product is WX2, an in-memory analytical database that is said to have got round the performance constraints of former data warehousing architectures based on the likes of Ingres and Postgres, which are said to introduce bottlenecks caused by the need to read and write to disk. By running in memory instead, Kognitio makes the bold claim that WX2 is the industry’s fastest and most scalable analytical database.
Pablo 1.0 is an extension to WX2 that creates in-memory images of an organisation’s database or data warehouse. By combining these images with a metadata layer and MDX connector, ad-hoc virtual cubes can be created and queried as often as required. The firm says this enables users to perform high performance analytics on high volumes of data, so-called "extreme analytics".
Speaking to CBR, Kognitio CEO Roger Llewellyn said that the key to WX2 is not just that it runs in memory but that it is also massively parallel. "Running in memory is great, but if you are not also massively parallel then the best performance you can get is from the fastest memory in a server," he said. "With our massively parallel in-memory architecture you can add more servers and so get an incredibly fast and scalable analytics database."
WX2 though does not create OLAP cubes like other warehouse architectures, yet many front-end query and analysis tools still need to suck in OLAP cubes to work. Pablo allows WX2 to get round this drawback by providing virtual OLAP cubes that can be ingested by the likes of IBM Cognos or SAP BusinessObjects. And thanks to the fact that Pablo ships with another tool called A La Carte which adds features to Microsoft Excel, Excel users will be able to create and investigate virtual OLAP cubes containing tens of Terabytes of data.
Also, says Llewellyn, because the OLAP cubes are virtual they can be created in a matter of seconds instead of in hours. That means that users can create new cubes pretty much as often as they like, and in so doing use their trusty BI tool of choice to query a vast data warehouse very rapidly (assuming it’s built on Kognitio WX2).
"While powerful, the potential of OLAP cubes has always been constrained by the time, tools and expertise needed to build, populate and use them," said Llewellyn. "With Pablo, these constraints are eliminated: virtual cubes can be built, populated and queried on the fly, with the minimum of time before data is ready to be examined. At the same time, by allowing users to take advantage of the familiar environment of Microsoft Excel, there is no need for extensive training before organisations can take advantage of this data. While Pablo does not immediately mean the end of physical cubes, the ability to virtualise large and complex cubes is an important leap in the evolution of OLAP."
Kognitio is based in Bracknell, UK, with North American headquarters in Chicago.