Should you use tabular or multi-dimensional SSAS mode? Here's the answer!
Part six of a seven-part series of blogs

If you're building a data warehouse, you'll want to know whether you should be creating cubes using the legacy multi-dimensional Analysis Services server mode, or creating data models in the new tabular mode. This blog gives Wise Owl's take on the subject.

  1. Analysis Services - Tabular or Multi-Dimensional Model?
  2. What the two SSAS Server modes have in common
  3. Creating a tabular model
  4. Creating and using a multi-dimensional model cube
  5. Comparing tabular model DAX with multi-dimensional MDX
  6. Other differences between multi-dimensional and tabular (this blog)
  7. Conclusions and a recommendation

This blog is part of our online SSAS Tabular tutorial; we also offer lots of other Analysis Services training resources.

Posted by Andy Brown on 07 November 2015

You need a minimum screen resolution of about 700 pixels width to see our blogs. This is because they contain diagrams and tables which would not be viewable easily on a mobile phone or small laptop. Please use a larger tablet, notebook or desktop computer, or change your screen resolution settings.

Other differences between multi-dimensional and tabular

This blog has argued strongly that you should choose the tabular server mode for SSAS, as it is so much easier to understand and use (and is integrated so much more closely with Excel and PowerPivot).

Above all, tabular models build on relational database theory, whereas to use multi-dimensional models you'll have to get your head round a completely new way of working.

However, in the interests of full disclosure, you should be aware of the following limitations of tabular models:

  • You can't define role-playing dimensions.
  • You can't define many-to-many relationships.
  • You can't create parent-child hierarchies.
  • You can't define actions (for drill-downs).
  • Write-back is not supported.
  • You can't use custom roll-ups.

This isn't a full list, but in general the extra functionality provided by multi-dimensional models is either obscure and not particularly useful, can be modelled around using tabular or is likely to be included in future tabular releases. 

 

This blog has 0 threads Add post