Introducing Optimize CPUs for Amazon RDS for Oracle

Posted on: Jun 21, 2018

(Updated) Amazon Relational Database Service (RDS) for Oracle offers a new capability called Optimize CPUs. It provides two ways to enhance the value of your Oracle Database licenses: you can specify a custom number of cores when launching new instances, and you can disable Intel Hyper-Threading (HT) Technology.

Optimize CPUs for RDS Oracle utilizes the recently announced Optimize CPUs for Amazon EC2. Together with the recent introduction of X1 and X1e instances, you have multiple ways to manage your database license utilization.

Many Oracle Database workloads need high memory, storage, and I/O bandwidth, but are not CPU-bound, so you can reduce the number of vCPUs in your deployment without affecting performance. For example, the X1e.8xlarge instance type offers 32 vCPUs but you can choose to launch it with 4, 8, 12, 16, 20, 24, 28, or 32 vCPUs. CPU-optimized instances have the same price per hour as full-sized instances, but the ability to manage the number of vCPUs enables you to optimize your vCPU-based license costs according to your own needs.

Intel HT Technology is enabled by default on Amazon RDS for Oracle instance types, but some applications aren’t designed to benefit from it. Compute-intensive workloads may experience performance degradation due to context switching, and perform better with single-threaded CPUs. By disabling Intel HT Technology, you can fine tune application performance and license utilization if you're a Bring Your Own License (BYOL) database customer.

You can configure the number of vCPUs and disable Intel HT Technology via the RDS Management Console, AWS SDK, or CLI. For further details, read our blog, and visit the documentation pages for Configuring the Processor of the DB Instance Class.

Amazon RDS for Oracle makes it easy to set up, operate, and scale Oracle Database deployments in the cloud. Optimize CPUs is now available in all public RDS regions; see RDS Oracle Pricing for detailed availability.