Organizations that have only explored or just started using data and analytics as well as reporting solutions such as Power BI at might need to consider what cloud platform they’ll use to store their data models in the future. Enterprises have two options: Power BI Premium and Azure Analysis Services (AAS).
Power BI Premium Capacity allows users to create a Premium workspace, which can be used for sharing Power BI reports on Power BI portal across the organization for users on free license. With Azure Analysis Services, users get an analytical model, but there’s still a need to additionally purchase either Power BI Premium Per User or Premium Capacity (i.e., for sharing reports). Note that it’s different with Power BI Premium Per User, as developers can share reports only for or with other users who have Power BI Premium Per User license.
Power BI Premium Gen2 and Azure Analysis Services have their own advantages, drawbacks, and caveats depending on how they can address the organization’s unique business needs. In our guide, we take a more in-depth look at these two platforms based on how our own clients decided to use one over the other, such as license modeling, costs, and security.
License model
In Power BI Premium, companies pay for license:
-
Per user: Monthly fee depending on the number of users
-
Per capacity: Monthly fee for Premium capacity (i.e., CPU usage and data model size limit), depending on the business’s needs. The capacity might be automatically scaled up and down (a feature for additional fee) and reports can be distributed to users with free license.
In Azure Analysis Services, companies pay for a fixed capacity (i.e., in CPU and server memory usage), but not for Power BI license. It’s possible to automatically scale the capacity up and down with other Azure services.
Cost comparison
The cost of using Azure Analysis Services to serve a large data model is lower in most cases compared with Power BI Premium Capacity. Note that the cost of Power BI Pro licenses required for report development should be added to every option as well (this will have the same fixed cost). It’s also worth mentioning that the capacity, based on the company’s needs, depends not on the size of the organization itself, but the level of Power BI adoption and analytical systems used.
Large companies that work with flat files (e.g., .xlsx, .csv) would consume less gigabytes than small organizations that work on complex source systems and processes, which use a lot of memory. Azure Analysis Services would be costlier here as it has a limit of 25 GB of memory. It would still require buying a Power BI Premium Capacity (at least P1 Tier), which could serve the data model in that limit in its price.
Memory usage and data volume
Azure Analysis Services has scale limits per server, while Power BI Premium’s are per model. When working with multiple data models, Power BI Gen2 can be a less expensive choice.
Azure Analysis Services might be a viable option if the company builds large data models, but it should still be combined with the Power BI Premium Capacity option to enable self-service reporting. Moreover, only about 70% of memory could be dedicated to store data model in Azure Analysis Services, and the rest needs to be treated as a buffer for query processing and user activities when memory in Power BI Premium Capacity is fully utilized.
Query processing performance
Power BI is faster than Azure Analysis Services. Power BI Premium’s architecture is less affected by overall load, temporal picks, and high concurrency, which enables processing to be managed more smoothly.
Moreover, Power BI Gen2 has the advantage of having more incremental features available. When working with very large datasets and there’s a need to refresh or add data from the most recent period, Power BI Premium will be the best choice. Its refreshes are faster and more reliable, which improve the decision-making process. Azure Analysis Services does not have this capability yet, and the incremental-like refresh can be implemented through custom workarounds and isn’t a straightforward automated process.
Security and data protection standards
There are some differences in Microsoft’s data protection standards, and if the organization has very strict data governance and protection requirements, Power BI Premium might fall short on them. Although this cloud solution does not have a firewall feature, it offers exceptional network security model including Azure Virtual Network (VNet) and Private Links.
Contrary to Power BI Premium, Azure Analysis Services offers an access to Dynamic Management Views (DMVs) and Content Level Security settings from which security can be controlled with more granularity. This solution gives the possibility to implement additional settings such detailed logging and changelogs. When it comes to network security, firewall needs to be set up and IP addresses for computers accessing the server should be configured.
Additional features, learning curve, and tool integration
Power BI Premium Gen2 is a superset of the functionalities of Azure Analysis Services. It has many powerful premium components such as dataflows, datamarts, deployment pipelines, and paginated reports, which make the solution more competitive against Azure Analysis Services. Microsoft works constantly in developing and adding new features so stayed tuned for more features in Power BI.
Migration to Power BI Premium enables companies to centralize and standardize all business intelligence assets, which can make it easy for Power BI developers to learn from and work with. Azure Analysis Services requires different skills since this environment has different configurations and maintenance processes. Integrating tools as well as streamlining resources and organizational structure might be fundamental factors for selecting which platform to invest in.
Download our guide, “Power BI Premium Gen2 vs. Azure Analysis Services: Which Is Better?” to have a more comprehensive comparison of both platforms, including case studies and in-depth comparisons based on some of real-life projects we’ve worked on for Fortune 500 and global companies. We tackle the various aspects that technology leaders, decision-makers, and business teams should consider before choosing to use either platform, including options for infrastructure setup. Our guide also delves into their query processing performance and memory management as well as analyze each platform’s additional features and learning curve.
