AWS announces Amazon Aurora PostgreSQL Optimized Reads

AWS announces Amazon Aurora Optimized Reads for Aurora PostgreSQL, a new price-performance capability available on new r6gd and r6id instances that delivers up to 8x improved query latency and up to 30% cost savings compared to instances without it, for applications with large datasets that exceed the memory capacity of a database instance.

Optimized Reads uses the local NVMe-based SSD block-level storage available on r6gd and r6id instances to store ephemeral data, reducing the data accesses to/from network-based storage offering improved read latency and throughput. These instances host temporary tables on the local storage (instead of network-based storage), delivering improved query performance for complex queries and faster index rebuild operations. Optimized Reads instances using I/O-Optimized use the local storage to extend their caching capacity. Database pages that are evicted from the in-memory buffer cache are cached onto local storage to speed subsequent retrievals of that data. This delivers up to 8x improved query latency, and enables Amazon Aurora PostgreSQL Optimized Reads with pgvector to increase queries per second for vector search by up to 9x in workloads that exceed available instance memory, accelerating the performance of Machine Learning and Generative AI applications.

Customers can get started with Optimized Reads through the AWS Management Console, CLI, and SDK by modifying their Aurora database clusters or creating a new one using r6gd or r6id instances. Optimized Reads is available for Aurora Postgres 14.9 and 15.4. For more information visit our pricing page, and documentation.