A deep dive into the general availability of Oracle Database@AWS

In an exclusive conversation with DataQuest, Kambiz Aghili, Senior Vice President, Oracle Cloud Infrastructure, discusses the strategic significance of the Oracle Database@AWS General Availability, its impact on cloud migration, modernisation, and AI-driven innovation. The following is the full transcript of our discussion:

Could you start by briefly sharing the background of the latest announcement on the Oracle and AWS partnership?

We will be announcing the general availability of Oracle Database on AWS. As you are aware, last September, we announced the private preview. That was one single AZ. The services were not fully there, but it was in the spirit of the partnership. We wanted to announce something, and we had over a dozen private preview customers that we had hand-selected to manually help them onboard to the service. These include the top financial and telecom operators worldwide. The sentiment was amazing, with many of whom were moving from on-premises to the cloud. They were getting anywhere from 25 to 30 per cent performance improvements being in the cloud, as well as meeting all the compliance and regulatory requirements that were very important to their operations.

We are announcing that we’re formally GA-ing the service and Exadata Autonomous in two regions in the United States, US East and US West. Each of the regions we launch will carry two availability zones, two AZs each. We have also reached an agreement with AWS to launch into 20 additional regions beyond that in the coming 12 to 18 months, allowing us to expand globally. That includes Hyderabad, Mumbai, and a variety of regions in Latin America, Europe, Japan, and more. This would allow the service to have further global reach.

All our database versions that are going to be available under this include 19c and 23ai. Our latest Exadata hardware technology, X11M, is what these services will be running over. This is the same hardware, software, and OCI stack running physically from inside AWS data centres or the AWS availability zones that we are announcing, providing proximity co-location of the database and applications, and our native integration with their services.

This is going to be huge for both companies. We have seen the success of that with Azure and GCP, and continuing on that path on behalf of the customers to help them move to the cloud is what we are announcing tomorrow morning.

Could you also expand on the strategic significance of the Oracle and AWS partnership, particularly with this general availability?

The significance can be addressed in two ways. Based on the latest analyst data from Gartner and IDC, 80% of mission-critical workloads are still running on-premises. There are many reasons why they arrived there over the past 20 to 30 years, different lines of business, M&A and acquisitions, each with their technology strategy. Some were in Azure, some in AWS, some in GCP. Running Oracle workloads on VMs or infrastructure in those third-party clouds meant they were not running efficiently. They were not getting the best benefits of Exadata hardware natively, which provides best-in-class throughput and IOPS.

This is the first time that customers who love AWS and love Oracle technology will have access to the native Oracle hardware and software operating natively from inside AWS data centres. They will have the Oracle technology they have always loved. Eight or nine out of ten customers in financial operations and banking have been using Oracle for 20 to 30 years. Not only will they have access to that best-of-breed wherever they want it, but also, due to the work we’ve done on unified security, governance, access controls, and networking, they can have these services operate with one another across platforms with unified governance.

Instead of having to manage the AWS fleet separately from the Oracle fleet or handling security and access control separately, we are making these services natively available to operate together. This allows them to develop new applications with AI utilising the Oracle estates and technologies, integrated natively.

How does Oracle Database@AWS simplify cloud migration for enterprises?

Refactoring has always been a blocker in moving to the cloud. Customers running mission-critical workloads on-premises have built their own reference and solution architectures, manually connecting to different AZs. Moving these to the cloud has been daunting for them, how to map this into a cloud environment.

We provide several capabilities to make this easier. One is migration technologies made natively available. Customers can leverage Oracle Zero Downtime Migration, which supports both physical and logical migration models. They have access to Oracle Data Guard natively. As a result, they can move from on-premises to the cloud with ease.

We also provide surrounding services like CSS services to make migration easy. We publish reference architectures to make the mapping from on-premises to cloud straightforward, alongside various automations not only around data and operational governance but also around the migration process. We have integration with AWS CloudFormation, observability tools, and Terraform, all designed to automate the migration and post-migration tasks. This creates a very easy landing spot for customers moving to the cloud.

How does this offering provide flexibility and choice in a multi-cloud environment?

Customers are already in AWS CloudFormation, Oracle Estates, and OCI. The choice has always been there. But the ability to natively connect these different deployment models is what was missing. Customers can now choose to operate Oracle technology on a third-party cloud while running it on the native Oracle hardware stack, `something that never existed before.

We have also introduced unified billing. Customers no longer have to deal with different billings from both companies. AWS users can find Oracle Database on AWS directly in the AWS Marketplace as a first-class citizen. They can utilise AWS commitments and credits to pay for Oracle Database on AWS. There is full license mobility, BYOL is supported, and customers can carry over their support rewards.

This seamless experience, integration, unified billing, and marketplace presence make the choice real. Customers can run on native hardware with ultimate performance, co-located next to AWS’s hardware, ensuring minimum latency between applications and databases across regions. This proximity is something that didn’t exist before.

How does Oracle Database support application modernisation?

We have invested significantly in Converged Database, which gives customers and developers the ability to handle multi-modal data without dealing with different purpose-built data stores. Whether it's document, JSON, graph, relational, data warehousing, OLAP, or OLTP, Converged Database offers a unified interface for all data types.

This simplifies application development and modernisation. We also offer MongoDB API compatibility, so customers can lift and shift applications natively to Oracle without complexity. Our native integrations allow customers to run third-party services like Amazon Bedrock, SageMaker, and Redshift with zero ETL integration, seamlessly utilising Oracle alongside.

For modernisation, customers have access to a richer set of services across both clouds that work natively together, supported by the Converged Database, 23ai features, and built-in vector capabilities. This provides best-in-class in-database convergence and AI capabilities that integrate with AI models of their choice.

Since you mentioned AI, how are the AWS and Oracle Data AI capabilities helping enterprises in driving innovation? What role does Oracle Database play here?

We look at AI across several layers of the stack. At the infrastructure level, we have best-in-class global infrastructure, including superclusters with RDMA networking, block, object, file, and HPC storage systems.

Above that, we have Oracle Database AI and Vector capabilities built into the database, from autonomous managed databases to data science, ML, and HeatWave AutoML, along with data labelling tools. This suite allows customers to derive insights from data in unique ways.

All our applications, NetSuite, Fusion, and industry solutions, are infused with AI built into the database. For example, natural language interfaces enable customers to query data conversationally, and we return the relevant results directly.

Oracle Database@AWS further enriches this by allowing customers access to various models and software. On Oracle Database at Azure, we have native OpenAI integration. On Google, we have Vertex AI and Gemini Foundation model integrations. On AWS, zero ETL integration connects with Redshift, Bedrock, SageMaker, and more.

Our broad strategy is consistent: to make Oracle technology available wherever customers need it, in any cloud, with native service and deployment model integrations. This makes AI adoption simpler, handling different data types and model choices, offering a complete end-to-end AI stack.

How does this launch align with Oracle’s broader AI vision?

We embed AI throughout the entire technology stack. Our cloud infrastructure is constantly optimised for AI workloads like LLMs. We aim to provide customers with simple, elastic access to AI infrastructure, software, database, and application modernisation, end-to-end.