Amazon Aurora serverless: Up to 30% better performance, smarter scaling, and still scales to zero

Amazon Web Services announced significant performance and scaling improvements to Aurora serverless, its autoscaling database service that can scale from zero to enterprise workloads automatically. The updated platform version 4 delivers up to 30% better performance than previous versions and introduces an enhanced scaling algorithm designed to handle workloads where multiple tasks compete for resources, such as busy web applications and API services. The improvements are particularly optimized for agentic AI applications, which typically exhibit unpredictable usage patterns with bursts of activity followed by long idle periods. Aurora serverless automatically adjusts capacity to match these workload patterns while maintaining its core capability to scale down to zero during periods of inactivity to minimize costs. All new Aurora serverless clusters will automatically launch on platform version 4, while existing clusters on older platform versions can upgrade through pending maintenance actions, cluster restarts, or blue/green deployments.