Metadata contention in tempdb has historically been a bottleneck to scalability for some workloads running on SQL Server. SQL Server 2019 introduces a feature named memory-optimized tempdb metadata which effectively removes metadata contention for tempdb-heavy workloads. In this session, we will understand how this feature works, how it is implemented, and when it is beneficial to use. We will also compare it to other techniques such as in-memory OLTP to decide which approach performs best for a given workload.


Torsten (MVP, MCSE, MCT) has been working with Microsoft SQL Server since the turn of the millennium and is specialized in performance optimization for large SQL Server environments.

He has been with bp as a database engineer for over two decades and collaborates closely with Sarpedon Quality Lab as a Principal Microsoft SQL Server Consultant. Torsten is the founder of and has given more than 170 presentations on Microsoft SQL Server performance optimization at national and international conferences and user groups.

Torsten has been recognized as a Microsoft Most Valuable Professional for Data Platform since 2018.