Does Microsoft Fabric replace Power BI? and other Questions…

Questions and Answers on Microsoft Fabric
  • Does Microsoft Fabric replace Power BI?
  • Is Power BI being rebranded as Fabric?
  • Is Power BI now Microsoft Fabric?
  • Is Microsoft Fabric free the same as Power BI?
  • Is Microsoft Fabric worth it?
  • Why do we need Microsoft Fabric?
  • Is Microsoft fabric the future?
  • Will Microsoft fabric replace Azure?
  • What is Microsoft fabric vs Databricks?
  • How many companies are using Microsoft Fabric?

We’re seeing a lot of questions from our customers about the relationship between Power BI and the new Microsoft Fabric.

Here we’ll give a short overview of the relationship between the two popular data tools./ Then we’ll answer some of these specific questions.

First an Overview of Microsoft Fabric

Fabric is the new suite of data tools that Microsoft has recently launched, and is rapidly developing. It will provide the basis for all business analytics and reporting, predictive analytics, machine learning and business intelligence.

It brings all your data together into a “One-Lake” data lake where you can efficiently clean and automate it. By bringing these multiple data sources together you can create real business intelligence.

Power BI is now a part of Microsoft Fabric, but does Microsoft Fabric replace Power BI? – No.

Power BI is where, after all the data gathering and preparation is complete, we are most likely to create our reports and distributions.  Power BI is effectively our main output for the data we have gathered, massaged, cleaned and modelled in Microsoft Fabric.

You can still continue to use Power BI exactly as you have been doing for the last 5 years. There’s no indication of existing features being switched off, though there are some features duplicated or improved which Microsoft may start to push users towards.   But Fabric does deal with some of the obvious issues that Power BI users are known to experience.

These problems typically include:

  • Data is everywhere. It can be hard to get it from our diverse systems into Power BI to work on.
  • People get data into Power BI, but everyone does it differently.  So there’s massive duplication of effort, duplication of data (cost), and no single source of truth that has been properly prepared.  So person A might be reporting on Sales, and so may person B, but the answers can be different because they have imported and managed the data different ways.
  • Everyone forms their own data store, and silos abound.
  • Scalability has always been an issue with larger datasets.  You could only import so much data into Power BI before the file became too large and slow.  And if you use the “Shortcut” method to connect directly to data in its source system, there were often limitations to performance and functionality.  Now in Fabric the data is brought into “One-Lake” and rapidly accessible there. Even if that data is being mirrored from a separate system, to you it is local, fast, and you can use all the Fabric tooling to efficiently work on it.
  • ML and predictive analytics are limited in traditional Power BI. There are some nice AI-based visuals on reports, but the majority of the ML and predictive functionality simply wasn’t there. So you’re looking at the past, not the future.

Another element of scalability is the ability to use Azure power to rapidly manage big data.   Most predictive analytics is based on the “Data Science” experience in Fabric. Here they use big data tooling and methods including using Apache Spark to run any number of parallel processes.  These processes can scale up or down as required, so when you have a massive data processing or machine learning job to do, dozens of parallel processes can spin up in no time and process your data in just a few seconds.

You can see that as data sources get more diverse, data quantities grow, and the need for rapid business intelligence becomes ever more pressing, Fabric provides a new set of tools that were not previously available in just Power BI.

So to the Questions and Answers:

Q. Does Microsoft Fabric replace Power BI?

A. Power BI is still there, and can do what it always did.  But now it is also a part of Fabric, and has acquired these additional powers to manage more data, and better, when you need them.

If you ask, “Does Microsoft Fabric replace Power BI?”, the answer is a clear no. Certainly Fabric including Power BI is now a much more powerful set of tools, and addresses the challenges that users experienced in Power BI in the past. These challenges typically prevented companies delivering excellent Business Intelligence, or getting going with AI and Machine Learning in the past. It has been a priority of Microsoft to unlock these features and get more people using their data better.

Q. Is Power BI being rebranded as Fabric?

A. No. You can see that Fabric is now a lot more than Power BI, so this is very much more than just a rebrand.

Q. Is Power BI now Microsoft Fabric?

A. Also no, but Power BI is now a key component part of Microsoft Fabric.

Q. Is Microsoft Fabric free the same as Power BI?

A. Again not really.  There is a free trial of Microsoft Fabric, which includes Power BI and other Fabric features, usually for 60 days.   There is also still the free version of Power BI, which is the desktop installable version only. This allows you to import data on your local machine and create reports.  But there is very little sharing with the Power BI Desktop, none of the collaborative features (Workspaces) and none of the cloud power of Fabric.

Both free trials are an excellent way to get a feel for the product and deliver some early business value, but they are different in important ways.

Q. Is Microsoft Fabric worth it?

A. Well this depends on what you want to do.  

  • Firstly, Fabric is very scalable, so you can purchase very small amounts of processing power and run Fabric quite cheaply.
  • Secondly it scales massively and can process almost limitless amounts of data. This allows for collaboration between very large teams, so it should get you using data to generate intelligence in ways that were not previously available.
  • Microsoft make a compelling case that the cost of ownership of these resources in the cloud will be much lower than if you were to set up “On Premises” type servers in the traditional manner.   By the time you have bought the hardware, provided suitable space for it, hired a tech team capable of looking after it, and maintained and updated it from each year to the next, they claim that using managed cloud services will be far cheaper, and retain a level of scalability, flexibility, and inter-operability that you can’t match “On Prem”.
  • In terms of “is it worth it” each user needs to make their own decisions. To assist we can help you with the ROI calculations you may need to make a business case for it.

For a company wanting to make better use of its data, make more informed business decisions, start benefiting from predictive Analytics, and provide intelligence based on data from a diverse range of systems, we believe that the answer is likely to be a resounding “Yes”.

Q. What does Microsoft Fabric cost?

A. We’re seeing a lot of questions about this. With Fabric now being on a consumption model the answer takes some calculation – or experience to be sure. We do have some information on our Fabric pricing page.

What we can tell you about the costs of a BI project is:

Time & Expertise: The largest part of a project cost is typically the expertise and time required, as with all data products, to get them implemented and working correctly. Its an important part of Fabric that “Data Demoncritization” means that more people can productively work witth and access data. But this still requires that skilled professionals will be the best people to provision the data, and carry out the critical tasks of data engineering, data warehousing, data science, and data analysis, all of which will contribute towards the quality of the output that you need.

Cost Reductions: From this it follows that the improved tooling and AI functionality will save costs rather than increase them, if you have been doing your reporting and BI in the past using older tools. But there is still the migration work to be done, so your project will have to be planned and costed carefully.

Compute and Storage pricing: There is a lot of published data about the licenses and “capacity units” so your computing and storage costs can be calculated. We can help clients with this process. One high level pricing snippet from our pricing page includes:

  • Smallest Capacity Unit: The lowest level of paid subscription is the F2 (2 capacity Units), which at the time of writing costs GBP £252 per month.
  • Largest Capacity Unit: At the top level the F2048 (2048 capacity units) offers the capability of massive big data processing and LLM models, and costs slightly over GBP £250,000 per month.
  • Storage: There’s an additional Fabric storage charge of £0.02 (2p) per GB per month.

This information isn’t enough though to determine what an individual project is likely to need. There will either be a very detailed analysis project to get a reasonable estimate of costs, or you can start small and see how costs respned as you add data and functionality.

Of course with any project its critical that the return on investment is a key element of the Go / No Go decision, and we’re happy to work with you on putting this together.

Q. Why do we need Microsoft Fabric?

A. I’ll repeat the answer above.  Only to add that we need Fabric because:

  1. We all want to make better business decisions.
  2. The volumes of data available to us are growing rapidly, and we have to be able to manage if effectively.
  3. Security threats are continuing to expand and we have to be able to secure our data effectively.
  4. Costs of IT are high, so we need to get the right data into other users’ hands, so they can make best use of it, rather than leaving all data work, processing and reporting in IT.
  5. Technology advances in AI and ML mean that some of our competitors are starting to steal a march on us in intelligence and automation.  If they aren’t yet we’re concerned about when one of them does.  Fabric makes all these technologies available to us in an accessible manner, so there’s no longer a reason for any business to be left behind by its competitors in a technology race.

Q. Is Microsoft fabric the future?

A. Yes probably.  It is one of our strong options for the future.  Though not the only big data, Lake, reporting and BI solution out there, it is Microsoft’s big bet, integrates with everything that Microsoft (and almost anybody) does, and is a very polished product.  While it’s not the only future-facing BI tool out there, it could sensibly claim to being one of the best.  It may well be the future for many of us.

Q. Will Microsoft fabric replace Azure?

A. Definitely Not.  Azure is the “Hyperscaler” cloud platform on which Fabric, and many other applications run.  Azure is the direct equivalent of Amazon AWS or Google Cloud.  It is not an individual application.  Azure is one of the platforms (infrastructure) on which all other cloud applications run.  There’s no more equivalence between Fabric and Azure, than there is between, for example, a car and a road.

Q. What is Microsoft fabric vs Databricks?

A. That’s a question requiring detailed consideration.

Its no surprise that Microsoft have included Databricks within in Fabric – after all they invested a great deal of money in Databricks themselves.  Much Databricks terminology and technology has been co-opted into Fabric, a testment to the success of Databricks since 2013. The term “Lakehouse” and the extensive use of Apache Spark are the first two to mind (Databricks were actually part of the team that developed Apache Spark) now a key technology within Fabric. 

Microsoft Fabric - does Fabric replace Power BI? How does Fabric compare with Databricks?

V

Databricks - where are the overlaps between Databricks and Microsoft Fabric?

There are many DataBricks functions that you can call from within Fabric. For example you can run a Databricks data transformation in a Data Pipeline while bringing data into Fabric,  or you can connect to DataBricks data and then use it directly in Fabric.

There’s a lot in common between the two platforms, and a decision should be made with care on each individual circumstance.  Here are a few notable differences:

AreaDataBricksMicrosoft Fabric
HostingRuns on any cloud platform, Azure, AWS, Google Cloud or other.Runs on Azure but can gather data from anywhere.
Code vs.
No Code
Takes some code knowledge to become productiveLots of no-code and low-code features.
Transition from?Natural to use if you are existing Notebook and Apache Spark usersEasier to start if you are already a big user of Microsoft Power BI or SQL Server.
Established user baseEstablished user base since 2013New technology though it does cater to established Microsoft user bases PowerBI, SQL Server and Synapse.
CostBoth are consumption-based pricing models with multiple price and performance tiersSome tests show 30% lower processing costs in Fabric vs Databricks with slightly faster performance in big data processes.

In Summary: both Fabric and Databricks are highly effective big data and BI tools.  Picking the right one for you should be a very careful selection process with experienced support. Here’s a good article on Medium with more comparisons.

Q. How many companies are using Microsoft Fabric?

A. We don’t know yet. 

Microsoft will know but we’ve not seen anything from them to say. 

Fabric only went officially live (General Availability) in November 2023, so its very new.  But we can see a lot of large organizations engaging with it and publishing information about it.   A quick search shows  it in use in several global companies already including KPMG, Cap Gemini, Hitachi, Informatica, Sitecore, Dell, Accenture, and many more. 

There are hundreds of Microsoft partners (quite obviously) creating content about it, and use-case related questions coming from organizations small and large all over the globe.  

The indicators are that almost everyone oin the space is looking at it, and starting to plan how it will fit into their BI and Data future.  Some have already made a meaningful start, and others will become visible in the next few months. 

Microsoft have discribed Fabric as “Our biggest new data product since SQL Server”, which is almost ubiquitous and was launched back in 1989.  We are confident that Fabric will see similar uptake within the next 2-3 years.

Summary

Thanks for the questions everyone, and to our readers. We can see a lot of people are thinking about the way forward, and working our how they will improve their business intelligence, decision making, and use of machine learning and big data.

These are important questions for a future-facing business.

Please ask our team if you would like any help resolving them, we’re usually very happy to help with an exploratory session.

And we’re very happy to answer any similar questions – please send them in via one of our contact channels and we’ll happily provide our best answer.