This is a very significant day in the progression of the Microsoft Business Intelligence platform. Earlier this year, Power View was released with SQL Server 2012; a remarkable step forward in data visualization. It wasn’t all bad news that Power View only worked with the new tabular semantic models and PowerPivot worksheets published in SharePoint but it did limit our options. We’ve quietly lived with the fact that this great new reporting tool couldn’t be used with Microsoft’s flagship analytical data (OLAP) engine without building new semantic models. Well, now it does, and that’s very good news.
Microsoft has officially announced the availability of “Microsoft SQL Server 2012 With Power View For Multidimensional Models CTP”. T.K. Anand made the announcement on the Analysis Services & PowerPivot Blog that the CTP for this enhancement is now available to freely download from the Microsoft Download Center. Be mindful that this is a pre-released preview of a patch for SQL Server 2012 that should not be installed in a production environment. Further, there may not be an upgrade path for the CTP installation when the final patch is released to the market. It would be best to use a sandboxed test server or virtual machine snapshot so the RTM upgrade can be installed when it is available later on.
My understanding is that this is actually an enhancement to the Analysis Services query engine that enables DAX queries to interact directly with multidimensional models (what we used to call UDM or OLAP databases). Existing SSAS databases will need to be upgraded to the SSAS 2012 server.
3 thoughts on “Power View Now Works with Cubes!”
I have tested this out. Its not entirely true. You end up having to read the multi-dimensional cube into a tabular model. The DAX then runs against that tabular model. This is bogus. Microsoft should not be peddling such rubbish.
I’d be interested in the results of your testing in greater detail. I’m confused by the statement that you “end up having to read the multi-dimensional cube into a tabular model”. Please explain. My experience with this feature has been quite positive and hasn’t required any model re-work.