By Imad Birouty, Teradata Product Marketing Manager
In-memory database processing is a hot topic in the market today. It promises to bring high performance to OLTP and Data Warehouse environments. As such, many vendors are working hard to develop in-memory database technology.
Memory is fast, but still expensive when compared to disk storage. As such, it should be treated as a precious resource and used wisely for the best return on your investment.
Teradata Intelligent Memory does just that. Through advanced engineering techniques, the Teradata Database automatically places the most frequently accessed data in memory, delivering in-memory performance with the cost economics of disk storage. The 80/20 rule and proven real-world data warehouse usage patterns shows that a small percentage of the data accounts for the vast majority of data access. Teradata Database’s unique multi-temperature data management infrastructure makes it possible to leverage this and keep only the most frequently used data in memory to achieve in-memory performance for the entire database. This is cutting-edge technology and does not require a separate dedicated in-memory database to manage. And because it’s built into the Teradata Database, companies get the scalability, manageability, and robust features associated with the Teradata Database.
Forrester Research just released their inaugural Wave dedicated to in-memory: The Forrester Wave™: In-Memory Database Platforms, Q3 2015 evaluation, naming Teradata a leader. Teradata has always been a pioneer in scalable, disk-based, shared-nothing RDBMS. Because it has continued to evolve, change, and incorporate the latest technologies, the Teradata Database is now a leader in in-memory database processing too.
While the Forrester Wave evaluated Teradata Database 15.0., we are even more excited about Teradata Database 15.10 which utilizes even more advanced in-memory techniques that are integrated into the Teradata Database. New in-memory accelerators such as pipelining, vectorization, bulk qualification, and columnar storage are integrated into the Teradata Database and bring in-memory performance to all data in the warehouse, including multi-structured data types such as JSON and weblogs which are associated with Big Data.
We’ll be announcing availability of Teradata Database 15.10 in a few weeks, so look for that announcement.