PASS Summit is the largest conference about the Microsoft Data Platform technology. Started as a Professional Association for SQL Server community, now the PASS brand includes the many technologies included in the Microsoft Data Platform. It is a large and active community, and the PASS Summit is the biggest annual international event, but you will find hundreds of local and virtual events in SQL Saturday, PASS Local Groups and PASS Virtual Groups.

After two days of preconference (I delivered one about Data Modeling with Power BI), Wednesday is the start of the “large” conference. The first keynote was dedicated mainly to SQL Server – even if these days it is hard to understand where is the boundary between on-premises, cloud, relational and non-relational, Windows and Linux.

If you have already seen the news about SQL Server 2017, I suggest you to review the Rohan Kumar keynote. It is mostly a recap of all the new features introduced by this version, but there are some news also described in a fresh new blog post from Rohan itself. From my point of view the two most important news are:

  • In a few weeks we will have Microsoft SQL Operations Studio, a new light weight version of SQL Server Management Studio (I know, Microsoft will never use this comparison to avoid marketing overlapping message). Why yet another management tool? Because it will run on Windows, Linux, and Mac. Should I add anything else?
  • Improvements in SQL Data Warehouse performance introducing the new Compute-Optimized Tier (maybe less popular, but very important to those interested in this platform)

For Power BI and Analysis Services I expect some announcement this week (and – why not – a new version of Power BI Desktop?). However, in the main session delivered by Kamal Hathi and many other PMs from the Power BI team.

A nice demo was using Power BI with IoT data in a sport team. I worked on this many years ago, those days collectiong and transfering data for the analysis was the bottleneck. Now we literally have the opposite problem: selecting which data to analyze. Which is a much more interesting problem to solve.

However, the biggest impact is showing Power BI navigating over one trillion rows from Spark. Christian Wade made a show by slicing and dicing data using one trillion rows from a Spark database tracking location from social media applications, which included geocoding over time. This data is aggregated by the engine (Power BI? Analysis Services? Well, the engine is almost the same…) until a certain level and then, going at the detail level, Analysis Services runs the query to Spark (using DirectQuery? probably…)

I’ve been told that Power BI Premium will have this feature first, but I hope that more detail will be revealed in other sessions. I will keep this post updated. There are biggest changes ahead!