# What Is Microsoft Fabric? A Plain-English Guide for Business Leaders Author: Nikhil Bhansali S Author URL: https://maxwize.in/blog/author/nikhil-bhansali-s Published: 2026-08-26 Category: Microsoft Fabric & Power BI Category URL: https://maxwize.in/blog/category/microsoft-fabric-and-power-bi Meta Title: What Is Microsoft Fabric? Plain-English Guide (2026) Meta Description: What is Microsoft Fabric, what it is used for, what sits inside it, and what it costs. A jargon-free guide for business leaders, checked August 2026. Tags: Pricing, Guide Tag URLs: Pricing (https://maxwize.in/blog/tag/pricing), Guide (https://maxwize.in/blog/tag/guide) URL: https://maxwize.in/blog/what-is-microsoft-fabric # What Is Microsoft Fabric? A Plain-English Guide for Business Leaders Microsoft Fabric is one product that covers everything between your raw data and your reports: moving data in, storing it, cleaning it, querying it, and showing it in Power BI. Microsoft sells it as software as a service, so you rent compute by the hour rather than assembling five or six separate Azure services yourself. If your company already runs Power BI, Fabric is the layer that sits underneath and does the data work. ## What Fabric replaced A Microsoft data stack used to mean buying several things and wiring them together: Azure Data Factory to move data, Azure Data Lake Storage to hold files, Azure Synapse or Databricks to process them, a SQL database to serve results, and Power BI on top. Each piece had its own bill, its own permissions, and usually its own copy of the data. Fabric puts those jobs in one product with one storage layer, one permission model, and one bill. You do not size virtual machines, create resource groups, or choose storage redundancy settings. You buy a capacity, create workspaces inside it, and start building. Microsoft also publishes a migration path from Azure Synapse Analytics into Fabric, which is where the new features now land. That is the trade. You give up the freedom to mix best-of-breed tools, and you get far less plumbing to build and maintain. For a company with two data people rather than twenty, that trade usually pays. ## What is inside Microsoft Fabric Fabric is organised into workloads. Each one is a set of tools aimed at a job and a role, and they all read and write the same data. As of August 2026 the official list on Microsoft Learn is: WorkloadWhat it doesWho uses itPower BIReports, dashboards, semantic modelsBusiness users and analystsData FactoryIngest and transform data using Power Query and 200+ native connectors to cloud and on-premises sourcesData engineers, analystsData EngineeringApache Spark, notebooks and lakehouses for large or messy dataData engineersData WarehouseA T-SQL warehouse with compute separated from storage, storing data in open Delta Lake formatSQL developers, data architectsData ScienceBuild, train and deploy machine learning models, with experiment tracking and a model registryData scientistsReal-Time IntelligenceAnalyse data as it arrives: IoT sensor readings, application logs, clickstreamsApp developers, analystsDatabasesOperational SQL and Cosmos DB databases inside Fabric, plus Mirroring to replicate data from Azure SQL Database, Azure Cosmos DB, Azure Databricks and SnowflakeApp developers, DBAsIndustry SolutionsPrebuilt data models for specific sectorsSector teamsFabric IQ (preview)Shared business semantics: ontologies, Fabric Graph, data agentsData architects You do not have to use all of them. Most companies start with Data Factory plus Power BI, add a lakehouse or warehouse when the data outgrows a single semantic model, and leave the rest alone. If you are weighing up whether you need more than reporting, the [Fabric vs Power BI comparison](https://maxwize.in/blog/microsoft-fabric-vs-power-bi) covers that decision in detail. ## What Microsoft Fabric is used for The five jobs businesses buy it for: 1. **One version of the numbers.** Sales sits in a CRM, invoices in an accounting system, stock in a warehouse app. Fabric pulls all three into OneLake on a schedule and Power BI reports against the combined set, so revenue means one thing across the business. 2. **Reports that do not break at scale.** A Power BI semantic model refreshing 40 million rows from Excel and SharePoint eventually fails. Moving that data into a lakehouse or warehouse and pointing Power BI at it fixes the refresh problem for good. 3. **Retiring a manual reporting process.** Somebody downloads four exports every Monday and rebuilds a spreadsheet. That becomes a Data Factory pipeline and a dashboard, and Monday morning stops being a job. 4. **Live operational data.** Real-Time Intelligence handles machine telemetry, delivery tracking, order streams and logs, where an overnight refresh is not good enough. 5. **Forecasting and scoring.** Demand forecasts, churn scores and lead scoring built in Data Science and fed straight back into the same Power BI reports the sales team already opens. If your reporting problem is a modelling problem, Fabric will not fix it. Power BI Pro on its own handles a lot, and a well-built star schema repairs more dashboards than new infrastructure does. ## OneLake is the part that matters OneLake is the single storage layer beneath every workload. It is built on Azure Data Lake Storage Gen2, and each Microsoft Entra tenant gets exactly one, with one file-system namespace spanning users, regions and clouds. Inside it you create workspaces, and inside those, lakehouses, warehouses and other items. Two features of OneLake change how projects run: **Shortcuts.** A shortcut points at data that lives somewhere else and makes it readable in Fabric without copying it. Supported targets include Azure Data Lake Storage, Amazon S3 and Google Cloud Storage. Your finance data can stay in S3 and still appear in a Fabric report. **Mirroring.** Mirroring continuously replicates an existing database into OneLake. Azure SQL Database, Azure Cosmos DB, Azure Databricks, Snowflake and Fabric SQL database are supported. You get an analytics copy without writing an ETL pipeline for it. Both exist for the same reason: the expensive part of an analytics project is usually copying and reconciling data, not building the report. ## How Fabric is licensed Two things get licensed, and people mix them up constantly. **Capacity** is the shared compute pool. You buy an F SKU through Azure, sized in capacity units (CUs), from F2 (2 CUs) up to F8192 (8,192 CUs). Azure bills per second with a one-minute minimum and no commitment, and you can take a one-year or three-year reservation to cut the rate. Legacy Power BI Premium P SKUs still support Fabric, but Microsoft is retiring them and tells both new and existing customers to buy F SKUs instead. **Per-user licences** decide what each person may do: LicenceUS price (August 2026)What it allowsFabric (Free)USD 0Create and share non-Power BI Fabric items (lakehouses, notebooks, pipelines) in a workspace on an F or trial capacity. View Power BI content only on F64 or larger with a viewer role.Power BI ProUSD 14 per user per month, billed yearlyCreate, share and view Power BI content on any capacity size.Power BI Premium Per UserUSD 24 per user per month, billed yearlyPro rights plus most Premium features, including 48 refreshes a day and larger models. Does not provision a Fabric capacity. The number to remember is F64. On F64 or larger, anyone with a free licence and a viewer role can open a Power BI report. Below F64, every viewer needs a Pro or PPU licence. For a company with 300 report viewers, that rule alone can decide the capacity size, so read [Microsoft's licensing rules](https://learn.microsoft.com/en-us/fabric/enterprise/licenses) before you pick a size. ## What it costs Fabric capacity prices are regional. In the US East region, the pay-as-you-go rate is USD 0.18 per capacity unit per hour, checked against the Azure retail price list on 21 August 2026. That works out as follows: CapacityPay-as-you-go (US East)Approx. per month if always onOne-year reservation, approx. per monthF2USD 0.36 per hourUSD 263USD 156F8USD 1.44 per hourUSD 1,051USD 625F64USD 11.52 per hourUSD 8,410USD 5,003 A one-year reservation cuts about 40% off the pay-as-you-go rate, and a three-year reservation cuts more. Check the [official Fabric pricing page](https://azure.microsoft.com/en-us/pricing/details/microsoft-fabric/) for your own region before you budget, since rates in India, the UAE and Europe differ from US East. Two things bring the real bill down. F capacities can be paused, and billing stops while they are paused, so a development capacity used in office hours costs a fraction of the monthly figure. And OneLake storage is billed separately and cheaply, at USD 0.026 per GB per month for hot storage in US East. The [full Fabric pricing breakdown](https://maxwize.in/blog/microsoft-fabric-pricing) goes through capacity units, bursting and smoothing in detail. ## Who should buy Fabric, and who should wait Buy it if you have data in three or more systems that has to be combined, if Power BI refreshes are failing or slow, if somebody is hand-building reports every week, or if you are already on Microsoft 365 and Azure and want one vendor and one bill. Wait if all your data fits comfortably in one system with a decent reporting module, if you have fewer than ten report consumers and no data engineer, or if your current problem is that nobody agrees on definitions. Fabric does not fix an argument about what counts as a qualified lead. ## How to try it without spending anything Microsoft gives you a 60-day Fabric trial capacity. It is provisioned as either F4 (4 CUs) or F64 (64 CUs), includes 1 TB of OneLake storage, and comes with a Power BI individual trial if you do not already hold a PPU licence. Copilot, AI features such as data agents, and Private Link are not available on the trial, and it is for evaluation rather than production. When the trial ends, workspaces revert to Pro and non-Power BI items stay in OneLake for seven days, so you have a week to attach a paid capacity before they go. Use the 60 days on one real reporting problem end to end rather than sampling every workload. Load one source system, model it, build one report the business asked for. That tells you more about fit than any demo. Once you know what you are building, the [Fabric implementation guide](https://maxwize.in/blog/microsoft-fabric-implementation-guide) covers the sequence and what outside help costs. If you are already deciding where data should land, read [lakehouse vs warehouse in Fabric](https://maxwize.in/blog/microsoft-fabric-lakehouse-vs-warehouse) next. ## FAQ **Is Microsoft Fabric the same as Power BI?** No. Power BI is one workload inside Fabric, the reporting layer. Fabric adds the data movement, storage, warehousing, streaming and data science layers underneath it. You can keep using Power BI on its own without ever buying Fabric. **Do I need an Azure subscription to use Microsoft Fabric?** You need one to buy a paid F capacity, because Azure bills it. The 60-day trial capacity does not need an Azure subscription, only a Microsoft Entra account and a Fabric free licence. **What is the cheapest way to start with Microsoft Fabric?** Start the 60-day trial capacity, which gives you 4 or 64 capacity units and 1 TB of OneLake storage at no cost. After that, an F2 capacity is USD 0.36 per hour in US East, about USD 263 a month if you leave it running, and you can pause it when nobody is using it. **Does Microsoft Fabric replace my data warehouse?** It can. Fabric Data Warehouse is a full T-SQL warehouse that separates compute from storage and writes data in open Delta Lake format. Most teams move in stages, starting with reporting and one or two source systems. **Do all my users need a Power BI Pro licence with Fabric?** Only below F64. On an F64 capacity or larger, users with a free licence and a viewer role can view Power BI reports. Below F64, every person who views Power BI content needs Pro or Premium Per User, and anyone creating Power BI items needs Pro or PPU on any capacity size. ## Where Maxwize fits Maxwize builds Fabric and Power BI setups for small and mid-sized companies in India, the UAE, the US and the UK, usually starting with one capacity, one pipeline and the report the business is currently rebuilding by hand. If you want a second opinion on whether Fabric is the right spend for your data, see our [Microsoft Fabric services](https://maxwize.in/services/data-analytics/microsoft-fabric) or [get in touch](https://maxwize.in/contact). ## FAQs Q: Is Microsoft Fabric the same as Power BI? A: No. Power BI is one workload inside Fabric, the reporting layer. Fabric adds the data movement, storage, warehousing, streaming and data science layers underneath it. You can keep using Power BI on its own without ever buying Fabric. Q: Do I need an Azure subscription to use Microsoft Fabric? A: You need one to buy a paid F capacity, because Azure bills it. The 60-day trial capacity does not need an Azure subscription, only a Microsoft Entra account and a Fabric free licence. Q: What is the cheapest way to start with Microsoft Fabric? A: Start the 60-day trial capacity, which gives you 4 or 64 capacity units and 1 TB of OneLake storage at no cost. After that, an F2 capacity is USD 0.36 per hour in US East, about USD 263 a month if you leave it running, and you can pause it when nobody is using it. Q: Does Microsoft Fabric replace my data warehouse? A: It can. Fabric Data Warehouse is a full T-SQL warehouse that separates compute from storage and writes data in open Delta Lake format. Most teams move in stages, starting with reporting and one or two source systems. Q: Do all my users need a Power BI Pro licence with Fabric? A: Only below F64. On an F64 capacity or larger, users with a free licence and a viewer role can view Power BI reports. Below F64, every person who views Power BI content needs Pro or Premium Per User, and anyone creating Power BI items needs Pro or PPU on any capacity size. --- This blog is powered by Superblog. Visit https://superblog.ai to know more. ---