Accelerate Business Analytics with Oracle Database In-Memory and Persistent Memory

With introduction of In-Memory column store, Oracle Database In-Memory (DBIM) significantly improves the performance for analytic queries as well as mixed workloads. This session will examine the Oracle Database In-Memory Column Store Architecture and configuration, how the data is populated into IMCS . We will discuss how to improve analytics and query performance in large data warehouse database using DBIM features such as In-Memory Expressions and In-Memory Virtual Columns as well as some latest enhancements of DBIM such as In-Memory support for external tables and Vectorized Query Processing and Automatic In-Memory (AIM) introduced in Oracle database 23c.For many large data warehouse and business analytics applications, it is important to have a large system memory to leverage the benefits of the Oracle DBIM. This can present a real challenge due to the limited size of physical memory (RAM) in the database server as well as the cost of the RAM. This presentation will discuss the experience of using large multiple Terabytes memory (RAM) system, as well as using persistent memory as alternative that allow us to build a much larger In Memory database, which provides a solution to the challenge. Several uses cases and performance benefit studies of leveraging persistent memory for DBIM will be discussed in the presentation