AWS' cloud graph database product Amazon Neptune is generally available in both US East regions, US West (Oregon), and EU (Ireland), more coming later this year
Tom Krazit / GeekWire :
Context & Ripple Effects
Six months after Amazon debuted Neptune at re:Invent 2017 as a preview managed graph database service, AWS is declaring it production-ready across both US East regions, US West (Oregon), and EU (Ireland), with more regions promised later this year. The move lands the same week AWS took Kubernetes support to general availability with EKS following other cloud vendors, making this a deliberate mid-year harvest of last fall's announcements.
The playbook echoes Google Cloud Platform, which pushed three enterprise database services out of beta back in 2016 Cloud SQL, Bigtable, and Datastore — evidence that database breadth has become a land-grab category where being first to general availability in each model matters.
First-order effects
- Customers can now run production graph workloads on Neptune in four launch regions instead of waiting on preview access, with EU (Ireland) giving European deployments a local home from day one.
- AWS converts a re:Invent announcement into billable infrastructure within roughly six months, tightening the gap between preview and revenue.
Second-order effects
- Rival clouds face pressure to field their own managed graph offerings, since Neptune's arrival means graph joins relational, key-value, and wide-column as categories where AWS sells a first-party managed service.
- Each new Neptune region raises the switching cost for graph workloads already anchored in AWS, feeding the broader lock-in dynamic that region-by-region database rollouts create.
Third-order effects
- If the announce-at-re:Invent-then-GA-in-spring cadence holds — EKS followed the identical path — AWS's release calendar itself becomes a competitive instrument, letting it blanket every database category faster than slower-shipping rivals.
- The longer arc runs from today's regional managed databases toward serverless distributed engines like Aurora DSQL introduced in late 2024, suggesting graph may eventually be absorbed into the same fully abstracted model rather than remaining a standalone product.
The trend: Cloud providers are racing to convert every database paradigm into a fully managed, multi-region service, with announcement-to-GA speed becoming the differentiator.