Microsoft Power BI has grown rapidly since its launch in 2015, with the emergence of big data as a powerful tool for organisations of all sizes. A Gartner and Forrester leader in Business Intelligence, Microsoft have created an easy to use tool at a price point that makes Power BI no longer the preserve of the enterprise, enabling businesses of any size to bring together previously disparate data sources into beautiful and easy to consume interactive reports.
With insightful decision making now accepted to create a major strategic advantage, Microsoft have announced it is expanding its offering with the launch of Power BI Premium, due late 2nd quarter 2017.
Gartner Magic Quadrant for BI & Analytics – February 2017
Dedicated Resource & Capacity
All users of Power BI Free and Power BI Pro share the same set of resources – also known as “shared capacity”. To be able to guarantee a level of performance, a number of restrictions are placed on users, including a 1GB model size and the number of times users are able to run data refreshes per day.
This is where Power BI Premium comes in, introducing the concept of “dedicated capacity”, providing a reserved and exclusive use of resources/hardware, removing previous throttling, restrictions and limitations.
With dedicated capacity allowing businesses to purchase based on their particular set of resources, Power BI content is made available to users throughout the company (and further) – without having to purchase individual licences. Providing this capacity-based licensing opens up a number of possibilities across the business, enabling everyone to take advantage of and operate in a data-centric environment.
Although Power BI brings data insight to the masses, Microsoft also understand the importance of keeping it secure. Power BI Premium enables administrators to control the appropriate level of access and usage dependent on their needs, this can be group (e.g. Management team), departmental and individually set.
Greater Scale & Performance
Providing dedicated resource and capacity means consistent, controllable performance and scale, fully controlled by administrators – enabling businesses and users to work with larger data sets, with data at its most fresh, for better busines performance.
At launch Power BI Premium will offer 3 tiers of service:
On Premises Capabilities
Microsoft understand that different businesses have different requirements. Some are not ready to transition all data and BI architecture to the cloud and some businesses need to work in a hybrid environment, especially within highly regulated industries. To accommodate this Power BI Premium also allows businesses to connect to on-premises servers — the Power BI Report Server allows sensitive data and reports to be kept behind the firewall.
Additional node sizes will be available in the future to support data model sizes in the hundreds of gigabytes and tens of processor virtual cores.
Along with improved user controls and capacity planning, Premium also introduces Power BI apps, which organise the benefits of the platform neatly, and allows users to explore tools for different insights “at enterprise scale”. Microsoft says: “Business users can easily install these apps and navigate them with ease, centralising content in one place and updating automatically.”
Coming Soon to Power BI Premium
Power BI’s features have grown rapidly since it was launched and the development team is not slowing down! In the upcoming months, following Premium’s launch a number of new capabilities will become available, including:
Larger data sets
The dataset limitation is currently set to 1GB. However, within the coming months the Power BI team will be removing this limitation for Premium customers, enabling them to build data models as large as the dedicated capacity can store. New SKUs will also offer larger capacity sizes allowing even bigger data sets.
With more data sources and larger data sets, refreshing becomes very system intensive. To address this Power BI Premium will give admin users the ability to set refresh policies and Power BI will incrementally refresh the data so only the newest data is loaded.
Pinning Datasets to Memory
Currently Power BI dynamically manages memory use to enable the most effective data transfer. However, there are some circumstances where users might want to override this by pinning datasets to an allocated amount of memory.
The Power BI system dynamically manages the memory use of the backend cores. When datasets stand idle without any queries hitting them, Power BI will silently evict them from memory to make room for more popular datasets. This dynamic memory management
mechanism allows Power BI to host millions of datasets more effectively. However, there are situations when users may want to override this behaviour – for example, for reports powering the CEO’s dashboard!
Dedicated Data Refresh Nodes
Refreshing data consumes a significant amount of resource. This puts a heavy work load on the node and means administrators may need to schedule data refreshes out of core hours to ensure performance is kept high. Dedicated Data Refresh Nodes this will split out queries from data refreshing meaning performance is maintained at all times.
When deploying Business Intelligence applications to a large user base and dataset, “Read-Only Replicas” will enable administrators to increase and optimise the running of queries, by giving them the ability to load balance across nodes and give businesses the power to handle even the largest set of queries.
For businesses with international offices, “Geo-Replicas” enables their users around the world to connect to a copy of the dataset closest to them, ensuring optimal performance, without latency issues physical distance can cause. These capacities can be provisioned in any of the global datacenters where Power BI is available.
As with all product roadmaps, the above features are subject to change. However, there are also a number of additional features not mentioned above that are being worked on, many of which will be announced at the Microsoft Data Insights Summit, 12-13 June 2017 in Seattle.