Within the last week, Microsoft has announced general availability of two important Azure & HD Insight-related analytics innovations, and a private preview of a feature that connects Azure Stream Analytics with Power BI Streaming Datasets. Let’s take a look at each announcement in a bit more detail.
Faster, more reliable machine learning at scale
Available as a preview since March of this year, Microsoft R Server on HD Insight is a high-performance, high-availability platform for predictive modeling, machine learning, and statistical analysis. It makes large-scale parallel analytics and machine learning faster, more manageable and enables faster access to analysis results. The highlights (as described by Microsoft) include:
- Terabyte-scale machine learning that is 1,000x larger than in open source R
- Up to 50x faster performance using R Server for Apache Spark and optimized vector/math libraries
- Enterprise-grade security and support backed by a Microsoft SLA
- Run distributed parameter sweeps and simulations with existing R functions
- Support for Spark 2.0
- Hive and Parquet data sources support
- Support for Spark data frames
- Easy setup for fast results
This marks Microsoft’s next step in making machine learning and related analytics tasks faster and more reliable at scale. More details are available here.
Sailing on the Azure Data Lake
Also, continuing the development of Cortana Intelligence, Microsoft’s managed big data and advanced analytics platform, the Azure Data Lake is a hyper-scale repository for trillions of data files of any kind. Your benefit? The ability to run parallel data transformation and processing programs at scale using U-SQL, R, Python and .NET. And because it’s a managed service, it’s all available without having to look after any of the infrastructure yourself.
Get more details by watching this video.
Private Preview: Azure & Power BI Streaming Datasets
Azure Stream Analytics is a key technology for extracting real-time insights from devices, sensors, infrastructure, and applications. This makes it perfect for Internet of Things (IoT) applications such as remote monitoring, or capturing insight from mobile phones and vehicles. This data could then be used to, for example, make a marketing offer based on a customer’s location. Connecting Azure Streaming Analytics with Power BI Streaming Datasets enable these analyses to be reflected more quickly within the Power BI visualization interface. Given that the whole point of real-time analytics is to enable faster reaction to events or opportunities based on the latest data, this is an important new capability.
Participating in the preview is as easy as sending an email to ASAtoPBIStreaming@microsoft.com with your Azure subscription ID. For full details about the announcement, as well as instructions on how to apply for the preview and set up Azure Stream Analytics with Power BI Streaming Datasets, click here.