These days, every Microsoft conference is a data event, and Microsoft's latest Ignite virtual event, which kicks off today, is no exception. With enough data and ML announcements to go around, I'll cover news pertaining to Azure Machine Learning, Power BI and Azure Purview in this post. ZDNet comrade-in-data Tony Baer covers Ignite's NoSQL, Redis, Synapse and Azure Arc news in a separate story:
Also read: Microsoft Ignite Data and Analytics roundup: Platform extensions are the key theme
Azure Machine Learning learns multi-cloud
First off, Azure Machine Learning will soon be offered as a multi-cloud Azure Arc-enabled service. That means you can train Azure ML models on other public clouds' infrastructure, or even on-premises. All you need is a Kubernetes cluster, which many enterprises run in their own data centers and for which all three major cloud providers have managed services.
This capability will allow customers who have data on other clouds to train Azure ML models in situ, without the burden of data movement. That also means customers won't have to incur the data egress charges that often apply when data is migrated off any given public cloud. Customers interested in signing up for the Azure Arc enabled ML preview can do so here.
Power BI lets subscribers pay their premiums
Back in September of last year, I wrote about Power BI Premium Per User (PPU), a new option that provides access to the features and capacity improvements in Power BI Premium, without having to pony up the enterprise-wide subscription charges that would otherwise apply. The service has been in preview since that time and has been free, with users wondering what the cost would be when that phase of the preview ended.
Also read: Microsoft announces Power BI Teams integration, NLP and per-user Premium subscription
Today Microsoft is announcing that PPU will go to full General Availability (GA) on April 2nd and will cost $20/user/month all-in. That means current subscribers to Power BI Pro, which is $10/user/month, will have PPU available as a $10/month add-on. That's pretty attractive pricing considering it provides access to important Premium-only features, like bigger models, more storage, text/vision AI services, paginated reports and the ability to work with Power BI data models from any tool that can talk to SQL Server Analysis Services or Azure Analysis Services.
Meanwhile, the enterprise-wide Power BI Premium offering moves to a new preview release that includes important autoscale capabilities. This makes Power BI much more elastic, allowing additional server nodes to be added on a temporary and policy-driven basis, by pairing the Power BI subscription with an Azure subscription, through which the additional compute power can be provisioned.
Also read: Microsoft's Power BI Premium delivers enterprise-grade features and bulk discounts
An increasing Purview
In December, Microsoft announced a preview release of its new Azure Purview data governance service, which effectively supersedes Azure Data Catalog. Among other capabilities, Purview provides a very powerful, searchable data catalog service, which enhances data discoverability in an organization. But to make Purview aware of the assets in your data estate, you have to define data sources, and then scan them. When the preview launched, only Azure data services, Power BI and on-premises SQL Server and Teradata were supported as sources.
Also read: Microsoft introduces Azure Purview data catalog; announces GA of Synapse Analytics
Today, Microsoft is announcing Azure Synapse Analytics and Amazon Web Services S3, as well as on-prem Oracle Database and SAP ERP (ECC and S4/HANA) instances, as additionally supported data sources, in preview. In the case of Synapse, both dedicated and serverless SQL pools are supported, which means both data warehouse and data lake assets can come into Purview. Support for Amazon S3 is a multi-cloud data lake play, support for on-prem Oracle and SAP is a hybrid-cloud play, both of which are smart. Since a catalog merely curates connectivity to, and information about, data, Purview shouldn't limit its sense of the data estate to Azure data assets only.
Center of gravity
Bringing enterprise BI capabilities to more users and taking both Azure Machine Learning and Purview to hybrid/multi-cloud land means Microsoft can broaden its data service prowess. And with that very franchise being such an important weapon in the cloud wars, these announcements are critical to Microsoft's overall industry position. Even if any service can run on any cloud, Microsoft is clearly using data to generate a gravitational pull to Azure.