The Analytics Platform System (APS), which is a renaming of the Parallel Data Warehouse (PDW), has a lot of obvious benefits, which I discuss here.  For those of you who find your database is getting too big, or becoming too slow, or you need to integrate non-relational data, check out APS, Microsoft’s MPP solution.

But there are a lot of non-obvious benefits to using APS that I have listed below:

• APS is a hardware platform that removes roadblocks for future needs. Be proactive instead of reactive!
• Numerous capabilities (i.e. PolyBase, mixed workload support, scalability, HDInsight) to get you thinking about better ways to use your data
• Much quicker development time due to speed of execution
• Allows removal of ETL complexity (temp tables, aggregation tables, data marts, other band aids and workarounds)
• Permits elimination of SSAS cubes or switch to ROLAP mode (so real-time data and no cube processing time)
• Reduction or elimination of nightly maintenance windows. Instead use intra-day batch cycles
• Tuning, redesigning ETL, upgrading hardware (more memory, Fusion IO cards), etc., for SMP to get 20-50% improvement versus 20-50x improvement with MPP
• SMP is optimized for OLTP while APS is optimized for data warehouses
• Allows a clear path to do predictive analytics via tools like Azure ML by having the disk space and processing power
• Don’t think of it as just a solution for a large data warehouse but rather for any size warehouse that needs faster query performance
• Faster query performance allows for adding more parameters to reports that also can be run real-time (instead of at night with fixed parameters) so business users can ask more sophisticated questions and execute ad-hoc queries.  This will result in a lot more self-service reporting
• Have the space and performance to consolidate all your various data warehouses and data marts to one place
• Decrease infrastructure costs through consolidation: reduced server count, reduction of duplicate data, decreasing licenses, viewer ETL pieces, less infrastructure management costs (hardware, SQL Server, OS)