This post was authored by Bob Ward, Principal Architect, and Jamie Reding, Senior Program Manager and Performance Architect, Microsoft Database Systems Group.
SQL Server continues to be a proven leader in database performance for both analytic and OLTP workloads. SQL Server 2017 is fast, built-in with capabilities and features such as Columnstore indexes to accelerate analytic performance and Automatic Tuning and Adaptive Query Processing to keep your database application at peak speed.
Recently, Hewlett Packard Enterprise (HPE) announced a new world record TPC-H 10TB benchmark result¹ using SQL Server 2017 and Windows Server 2016 on their new DL580 Gen10 Server. This new amazing result at 1,479,748 Composite Query-per-Hour (QphH) was achieved at price/performance of .95 USD per QphH continuing to show SQL Server’s leadership in price and performance.
HPE also announced the first TPC-H 3TB result² on a 2-socket system using SQL Server 2017 and Windows Server 2016 with their DL380 Gen Server. They achieved a stellar 1,014,374 QphH on only 2-sockets. These results continue to show how powerful SQL Server can be to handle your analytic query workloads including data warehouses.
SQL Server also is a proven leader for OLTP workloads. Lenovo recently announced a new world-record TPC-E benchmark result³ using SQL Server 2017 and Windows Server 2016. This is now the #1 TPC-E result in both performance at 11,357 tpsE and price/performance at 98.83 USD per tpsE for systems with 4 sockets or more. This result was achieved on Lenovo’s ThinkSystem SR950 server using 4 sockets at 112 cores which represents a 25% performance gain from the previous 4 socket result with 16% more cores.
SQL Server 2017 is the world leader in TPC-E and TPC-H performance, price, and value and continues to demonstrate it is one of the fastest databases on the planet, in your cloud or ours.
- ¹ 10TB TPC-H non-clustered result as of November 9th, 2017.
- ² 3TB TPC-H non-clustered result as of November 9th, 2017.
- ³ TPC-E benchmark result as of November 9th, 2017.
Source:: Microsoft – SQL Server