Parallel partition processing available for Tabular Models in SQL Server 2016 Preview

One of the great new features that was actually added back in CTP 2.0 of SQL Server 2016 Preview is the ability to process partitions in a table in parallel for Tabular Models. Today I want to highlight this highly requested feature and show you what this might mean for your Tabular models. With parallel partition processing, the partitions of a single table will now be processed in parallel instead of sequentially, as is the case prior to SQL Server 2016. Partitions are used to manage data in a table. For example, a single sales table can be divided into partitions by month. Let's say you have 2 years of data. When you create a partition by month, you split your table into 24 partitions. This now enables you to refresh each individual partition if necessary. You can refresh just the partition for the current month to load the most recent data, so you don't have to refresh all the data. But often, data in the past changes as well. In this case, all the data in the table needs to be refreshed periodically (several times a day or once a night) to make sure all the data remains current.