Amazon DocumentDB (with MongoDB Compatibility) introduces NVMe-backed R6gd Instances, offering up to 7x faster query performance for applications with large datasets exceeding regular instance memory. These instances can expand the data caching capacity by up to 7 times compared to a regular instance.
NVMe-backed R6gd Instances leverage the local non-volatile memory express (NVMe)-based SSD storage available on r6gd instances to store ephemeral data, reducing network-based storage access, and improving read latency and throughput. Database pages evicted from the in-memory buffer cache are cached on local NVMe storage. Additionally, temporary space is hosted on local SSD instead of network storage, reducing the latency of queries involving sorts by up to 2x and accelerating resource-intensive operations like index builds.
Customers can get started with NVMe-backed R6gd Instances through the AWS Management Console, CLI, and SDK by modifying their Amazon DocumentDB database cluster or creating a new one using R6gd instances. NVMe-backed R6gd Instances are available for Amazon DocumentDB V5.0 on both Standard and IO-Optimized cluster storage configurations. For more information visit our pricing page and documentation .