Exploring MS Fabric: A Power BI View

Recently at the MS Build 2023 conference, Microsoft made an exciting announcement about the introduction of Microsoft Fabric. As a Power BI professional and enthusiast, I was particularly curios about how this new offering would impact Power BI, So, let’s dive in and explore what MS Fabric is all about!

Do note that this is my interpretation of MS Fabric and what it means for Power BI. As of now, all knowledge was directly from the MS documentation released and made available.

Overview of MS Fabric

When I first heard of MS Fabric and started looking into it, I interpreted it as an umbrella term encompassing various elements, and it very much reminded me of how Power BI combined Power Query, Power Pivot, and Power View back in 2015/16. With that said, MS Fabric is a Software-as-a-Service (SaaS) solution that brings together different components of Synapse, such as data integration, big data processing, data warehousing, data science, and real-time analytics. The unique aspect of MS Fabric is that it combines all these elements into a single low-code/no-code environment, just like Power BI.

In a nutshell, MS Fabric provides organisations with a unified SaaS platform that simplifies the entire data processing journey.

Microsoft Fabric - vision

Now, I can’t mention MS Fabric without bringing up, OneLake! So, OneLake serves as a single data lake for the entire organisation. Think of it like OneDrive, but for analytics data, allowing easy access and sharing of data in the cloud. With OneLake, organisations benefit from having:

  • One data lake for the entire organisation
  • One copy of data for use with multiple analytical engines

So, how many OneLake’ s can we have? Well, with every MS Fabric tenant, one OneLake is automatically provisioned, hence ensuring a centralised approach to data management. However, to cater to different business areas’ independent activities, MS Fabric introduces the concept of domains. Domains logically group together data relevant to specific groups or business needs. Each domain has its own admins, contributors, and workspaces, and this promotes a data mesh architecture. Domains also facilitate data discovery and enhance trustworthiness through the application of endorsements.

What does this mean for Power BI?

Now, let’s discuss the relevance of MS Fabric to Power BI. It’s important to note that Power BI itself remains the same. By this I mean we can still open Power BI Desktop, ingest data from various sources, create reports, publish to the Power BI Service, etc. However, with MS Fabric, Power BI gains some additional features. Also, from my understanding the Power BI tenant will eventually transition into an MS Fabric tenant since Power BI is now part of MS Fabric.

Additional Experiences in the Power BI Portal

If you’re anything like me, you might have been wondering about the additional experiences in the Power BI portal – soon to be known as the MS Fabric portal. In the past, accessing different experiences meant navigating through the Azure Portal. However, things have changed now. Users can simply log into the MS Fabric Portal, just like accessing Power BI, to access all these experiences. The MS Fabric SaaS platform integrates Power BI, Data Factory, Data Engineering (Synapse), Data Science (Synapse), Data Warehouse (Synapse), and Real-Time Analytics (Synapse). So, if you’re accustomed to using everything Synapse, this change will be more noticeable for you compared to those who have primarily used the Power BI Service.

As a result of this, there are additional use cases when accessing the Power BI portal. Workspaces, which previously stored only Power BI artifacts, now will hold much more for each of these experiences. Also, each experience will offer a unique persona experience for the relevant users.

New Storage Mode – DirectLake

As mentioned above, one of the notable advantages of MS Fabric is the ability to read data from OneLake. It’s important to call out that everything is stored in a consistent Delta format. Why is this important? Well, it makes it easy to pull data from OneLake into Power BI and with that MS introduced a new storage method (connectivity type) known as Direct Lake, which combines the best of Import and DirectQuery. With Direct Lake, we can achieve fast and real-time performance while avoiding the common challenges I am sure you have all come across. As of now, Direct Lake is in public preview for Fabric Lakehouses, and this same storage method will be made available through public preview soon but for Fabric Warehouses.

Promoting Actions through Data Activator

MS Fabric also includes Data Activator, a feature that enables users to take action through its alert and notification capabilities. I have not explored this much; however, this functionality helps users stay on top of data changes and facilitates actions. As you can imagine, this goes hand in hand with reporting, as deriving and taking action from our data is fundamental.

Co-Pilot – “A whole new way to work”

Another exciting addition to MS Fabric is Co-pilot, an AI feature that enhances all experiences across MS Fabric. For Power BI, Co-pilot enables users to accomplish three core tasks. Firstly, it allows for the creation of reports by providing specifications. Secondly, it simplifies the process of creating DAX measures by typing what is required. Lastly, Co-pilot empowers users to explore and analyse their data by asking questions to uncover valuable insights. When I think of this, it’s another step in the evolution of how we consume data, however, I am still to try this out as it’s currently in private preview.

From Dataset Hub to OneLake Hub

Using Power BI, you are probably familiar with the Dataset Hub. This helps democratise our datasets across the organisation by making them easily discoverable and available. With MS Fabric, this is now known as the OneLake Data Hub, from herewe can continue to easily locate data items across workspaces but now also with the inclusion of MS Fabric items.  In addition I mentioned the use of domains in the section above and and how these offer a convenient way to view data relevant to specific business areas within the OneLake Data Hub.

What about licensing?

The first thing to call out is MS has provided a 60-day Fabric trial for all users to trial the new Fabric features. Also, if you cannot see the additional experiences through the Fabric portal, it does need to be enabled by your Power BI Service administrator, which I am guessing is now the Fabric Service administrator.

If you have a Power BI Premium capacity, you will not need to purchase an F SKU separately to take advantage of using and sharing MS Fabric content. The below diagram from the MS documentation shows the equivalent F SKU depending on which P SKU capacity you currently have. So, if you have a P1 SKU, you have a F64 SKU.

However, like with previous versions of Power BI, a P SKU was required for free read-only access to Power BI reports. This is the same with F SKUs. Anything under an F64 will require all users to have a Pro license to consume content. Also, the Power BI Free User licenses have now changed to the Microsoft Fabric Free User licenses. I’ll be sure to post any other licensing information that becomes available.

If you want to learn more about applying OpenAI to your organisation, watch our webinar with our head of data science and Microsoft’s UK Lead for Data & AI, as they provide a practical view on how organisations can leverage Azure OpenAI technology to drive real business value. You can also find more information on  Azure OpenAI with our Azure OpenAI Envisioning Workshop