Microsoft BI Conference, Day One

Gemini, Madison, Kilimanjaro and More

I know when a company puts on a show like the Microsoft BI Conference, there are going to be announcements, and pre-announcements of new products and features. The bullets:

Project Gemini - A self-serve BI application, rooted in Excel and an in-memory datastore the primary goal of which is to allow user creation of SSAS databases. Databases that are then managable and not 'spread marts' or 'spreadsheet purgatory'. Donald Farmer actually demonstrated filtering and sorting 20 millon rows on a thousand dollar quad-core pc.

Project Madison - With the DATAllegro purchase, Microsoft is quickly integrating the newly aquired MPP capability into the Madison release targeted for the first half of 2010. Jesse Franklin demonstrated clean-cache queries against a 150 terabyte data store accross 24 instances of SQL Server. The response times were screaming.

Project Kilimanjoro - This will be a focused release of SQL Server, that I understand to be a re-architecture (read euthanasia) of the failed Report Builder functionality in SQL 2005. Instead, a set of task focused add-ins to Excel that will pave the way for component based report writing. No demo, but very exciting rhetoric.