Oracle and AWS are deepening their cloud database partnership as enterprises look for ways to bring artificial intelligence closer to the business data already sitting inside mission-critical Oracle systems.
The companies have signed an expanded long-term strategic collaboration agreement aimed at accelerating migrations to Oracle AI Database@AWS, which is now available across 22 AWS Regions.
Oracle also made Exadata Database Service on Exascale Infrastructure generally available through the service, extending Exadata capabilities to a wider range of organizations through a consumption-based infrastructure model.
The bigger story is increasingly familiar across enterprise AI: companies want access to advanced AI services without first rebuilding the systems containing decades of operational data.
Oracle Database Moves Deeper Into AWS
Oracle AI Database@AWS allows companies to operate Oracle database workloads inside AWS environments while connecting them with AWS analytics and AI services.
That now includes integrations with services such as Amazon Bedrock and Amazon SageMaker, along with zero-ETL integration with Amazon Redshift.
Oracle says application-to-database round-trip latency can reach as low as 165 microseconds using AWS High Performance Networking and updated deployment configurations.
Customers can also use Amazon S3 for Oracle-managed backups, Amazon CloudWatch for monitoring and Amazon EventBridge for event-driven applications.
The idea is to remove one of the architectural compromises enterprises have historically faced: choosing between keeping Oracle workloads close to their existing database infrastructure or moving data into another environment to take advantage of cloud-native AI and analytics.
Exadata Gets a More Flexible Model
The new Exadata Database Service on Exascale Infrastructure could broaden the addressable market.
Rather than requiring customers to provision dedicated database and storage servers, organizations can specify the compute and storage capacity they need while databases operate across pooled storage infrastructure.
Oracle says this provides Exadata performance and availability while improving the economics for smaller workloads and multi-region deployments.
Instant thin cloning also allows organizations to create database copies for development and testing without duplicating all of the underlying storage.
That could become particularly useful as developers build AI applications against production-like enterprise datasets.
AI Comes to the Data
The AI implications may ultimately be more important than the infrastructure announcement.
Oracle AI Database incorporates capabilities including AI Vector Search, while AWS provides a growing collection of generative AI, machine learning and analytics services.
Connecting those environments gives companies another route for applying AI to operational information without repeatedly copying data into separate platforms.
CJ Olive Young, Kobalt Music Group and Barcelona’s Metropolitan Transport Authority are among the organizations already using Oracle AI Database@AWS for business-critical workloads.
Oracle and AWS are also expanding the partner ecosystem around the platform. Eligible AWS and Oracle partners can resell the service through AWS Channel Partner Private Offers, while qualified managed service providers can manage deployments for customers.
The Bottom Line
Oracle and AWS are addressing one of the less glamorous but more consequential challenges in enterprise AI: getting AI close enough to the data that actually runs the business.
Companies have spent decades building applications around Oracle databases. Few want to rip those systems apart simply to participate in the generative AI boom.
Oracle AI Database@AWS offers another path. Enterprises can keep Oracle technology at the center of critical data operations while gaining access to AWS infrastructure, analytics and AI services around it.
The expansion to 22 AWS Regions and the introduction of a more flexible Exadata model suggest this partnership is moving beyond multicloud experimentation and becoming a serious modernization option for large Oracle customers.