Skip to main content

Create a Scalable Data Model for Enhanced BI

Headshot of article author The Power BI Team

Managing millions of rows of data from multiple data sources requires a great deal of planning. Making this information readily accessible to BI users is just one (extremely important) goal. The information itself has to be worth obtaining—so you need to design your model to build and manage connections between data from a variety of sources, as well as relationships between multiple tables. Your data model is the foundation for all your data analysis.

When creating your data model, you need to consider key goals:

  • The impact on speed of processing
  • How you can optimize memory usage and performance
  • Scalability when handling growing volumes of data and requests

That’s why we’re introducing a deep-dive Microsoft Ignite session on Scalable BI and Advanced Modeling with Microsoft SQL Server Analysis Services and PowerPivot. Created for enterprise developers, this session shows you how to define a scalable data model in Microsoft SQL Server Analysis Services or Power Pivot to meet these essential goals. And we’re taking it even further: the right definition of your data model will be especially useful if you want to take advantage of the self-service navigation, discovery, and prediction capabilities of Microsoft Power BI to share all of this helpful information.

We’ll discuss how to use common design patterns to solve issues, such as handling relationships at granularities other than primary key in a table. And you’ll get some suggestions about hardware selection, which is particularly important for maintaining the performance of large models.

This session will be of most interest to BI developers, particularly those beginning to plan or implement an analytics solution using a tabular model.

If you haven’t already signed up for Ignite—the gathering of some of the best minds in the business—you can register here, or check out more sessions here.