Did you know Aurora Serverless can now hit 12 ACUs in under a second?

Hey everyone, it's me! I've got some exciting AWS news for you today: database auto-scaling just got a lot faster, and I can't wait to tell you about it!

AWS What's New aws.amazon.com

What was announced?

According to AWS What's New, Amazon Aurora serverless scaling just got a big speed boost. When your workload suddenly ramps up, Aurora serverless now reaches up to 12 ACUs (Aurora Capacity Units) of initial capacity within a second, and keeps scaling up from there to as much as 256 ACUs as your workload grows. Once things quiet down, Aurora serverless automatically scales back down to zero, so you only pay for what you actually use.

The story so far

Aurora serverless already had a mechanism to auto-scale based on workload. But this announcement specifically calls out "higher initial capacity during scale-up events," which tells us the initial ramp-up capacity used to be lower than it is now. That could be a real bottleneck for workloads like AI agents, which tend to sit idle for long stretches and then suddenly need a burst of processing power all at once.

What changes

With this update, the initial capacity during scale-up is much higher, so sudden bursts can be handled right away. This is especially useful for agentic AI applications: agents typically idle for a while and then trigger a flurry of activity when they kick off. Aurora serverless now scales along with your agents automatically, so developers don't have to worry about capacity planning and only pay for what they use.

Dive Deep

This faster scaling is enabled by default on any Aurora serverless cluster running platform version 3 or 4, with no configuration changes required. If you're still on platform version 1 or 2, you can upgrade directly to the latest platform version 4 to get these benefits.

You can check your cluster's platform version in the AWS Management Console under the instance configuration section, or via the RDS API's ServerlessV2PlatformVersion parameter. For pricing and Region availability, AWS points you to the Amazon Aurora Pricing page, and for the full mechanics, check the Aurora serverless scaling documentation.

Wrap-up

• During scale-up, Aurora serverless now reaches up to 12 ACUs of initial capacity within a second, then keeps scaling up to 256 ACUs
• It automatically scales back down to zero when the workload finishes, so you only pay for what you use
• Enabled by default on platform version 3 / 4 clusters, no configuration changes needed
• Clusters on platform version 1 / 2 need to upgrade to platform version 4 to get this benefit
• Especially well suited for agentic AI applications with long idle periods and sudden bursts of activity

If you're running workloads that sit quiet most of the time but suddenly get busy, this update is really made for you!