What is Microsoft Power BI Premium?
Microsoft Power BI Premium provides resources dedicated to running the Power BI service for your organization. It gives you more dependable performance and enables larger data volumes. Premium also enables widespread distribution of content without requiring you to purchase per-user Pro licenses for content consumers. For purchasing information, see How to purchase Power BI Premium.
Premium capacity and shared capacity
You take advantage of Power BI Premium by assigning workspaces to a Premium capacity. Premium capacity is a dedicated resource for your organization. Workspaces not assigned to a premium capacity are in a shared capacity. With shared capacity, your workloads run on computational resources shared by other customers.
The following image shows the relationship between Premium capacity and shared capacity, using the Contoso organization as an example.
|(1) Items within a Premium capacity||
|(2) My workspace in shared capacity||
|(3) App workspaces in shared capacity||
In shared capacity, Power BI puts more limits on individual users to ensure quality of experience for all users. By default, your workspace is in a shared capacity, including your personal My workspace and App workspaces.
The following table provides a summary of the differences between shared capacity and Premium capacity:
|Shared capacity||Power BI Premium capacity|
|Isolation with dedicated hardware|
|Enterprise Distribution to all users|
|Apps and sharing|
|Embedded API and controls||1|
|Publish Power BI reports on-premises|
1 Future enhancements coming to Power BI Premium.
Premium capacity nodes
Power BI Premium is available in node configurations with different v-core capacities. For more information about specific SKU offerings and costs, see Power BI pricing. A cost calculator is also available. For information regarding embedded analytics capacity planning, see Planning a Power BI Enterprise Deployment whitepaper. To summarize:
P nodes can be used for embedded or service deployments.
EM nodes can be used for embedded deployments only. EM nodes do not have access to premium capabilities, such as sharing apps to users that don't have a Power BI Pro license.
|Capacity Node||Total v-cores
|Backend V-Cores 1||Frontend V-Cores 2||DirectQuery/live connection limits||Max concurrent refreshes||Availability|
|EM1 (month to month)||1 v-core||0.5 v-cores, 2.5-GB RAM||0.5 v-cores||3.75 per second||1||Available|
|EM2 (month to month)||2 v-cores||1 v-core, 5-GB RAM||1 v-core||7.5 per second||2||Available|
|EM3 (month to month)||4 v-cores||2 v-cores, 10-GB RAM||2 v-cores||3||Available|
|P1||8 v-cores||4 v-cores, 25-GB RAM||4 v-cores||30 per second||6||Available (month to month is also available)|
|P2||16 v-cores||8 v-cores, 50-GB RAM||8 v-cores||60 per second||12||Available|
|P3||32 v-cores||16 v-cores, 100-GB RAM||16 v-cores||120 per second||24||Available|
1: Frontend v-cores are responsible for the web service. For example, dashboard and report document management, access rights management, scheduling, APIs, uploads and downloads, and generally for everything that relates to the user experience.
2: Backend v-cores are responsible for the heavy lifting, like query processing, cache management, running R servers, data refresh, natural language processing, real-time feeds, and server-side rendering of reports and images. With the backend v-cores, a certain amount of memory is reserved as well. Having sufficient memory becomes especially important when dealing with large data models or with a large number of active datasets.
Workloads in Premium capacity
Think of a workload in Power BI as one of the many services you can expose to users. By default, capacities for Power BI Premium and Power BI Embedded support only the workload associated with running Power BI queries in the cloud.
We now offer preview support for two additional workloads: Paginated reports and Dataflows. You enable these workloads in the Power BI admin portal or through the Power BI REST API. You also set the maximum memory each workload can consume, so that you can control how the different workloads affect each other. For more information, see Configure workloads.
Default memory settings
The following tables show the default and minimum memory values, based on the different capacity nodes available. Memory is dynamically allocated to dataflows, but it is statically allocated to paginated reports. For more information, see the next section, Considerations for paginated reports.
Microsoft Office SKUs for software as a service (SaaS) scenarios
|Paginated reports||N/A||20% default; 10% minimum||20% default; 5% minimum||20% default; 2.5% minimum|
|Dataflows||20% default; 8% minimum||20% default; 4% minimum||20% default; 2% minimum||20% default; 1% minimum|
Microsoft Azure SKUs for platform as a service (PaaS) scenarios
|Paginated reports||N/A||N/A||N/A||20% default; 10% minimum||20% default; 5% minimum||20% default; 2.5% minimum|
|Dataflows||27% default; 27% minimum||20% default; 16% minimum||20% default; 8% minimum||20% default; 4% minimum||20% default; 2% minimum||20% default; 1% minimum|
Considerations for paginated reports
If using the paginated reports workload, keep in mind paginated reports allow you to run your own code when rendering a report (such as dynamically changing text color based on content). Given this fact, we secure Power BI Premium capacity by running paginated reports in a contained space within the capacity. We assign the maximum memory you specify to this space, whether or not the workload is active. If you use Power BI reports or dataflows in the same capacity, make sure you set memory low enough for paginated reports that it doesn't negatively affect the other workloads.
In rare circumstances, the paginated reports workload can become unavailable. In this case, the workload shows an error state in the admin portal, and users see timeouts for report rendering. To mitigate this issue, disable the workload then enable it again.
Power BI Report Server
Power BI Premium also includes the capability to run Power BI Report Server on-premises in your organization. To learn more, see Get started with Power BI Report Server.
More questions? Try asking the Power BI Community
We'd love to hear your thoughts. Choose the type you'd like to provide:
Our feedback system is built on GitHub Issues. Read more on our blog.