Oracle’s database family gets trendy AI makeover

Oracle has celebrated the general availability of its latest database upgrade by renaming it.

Out is the “c” lineage that started with the Oracle 12 database in 2013, in is “ai” as the 23c database teased in September last year gets a makeover for the tech trend of the moment.

Big Red announced the new name along with the database’s general availability, which, we should note, only applies in Oracle Cloud Infrastructure (OCI) on Oracle Exadata Database Service, Oracle Exadata Cloud@Customer and Oracle Base Database Service, as well as on Oracle Database@Azure. Commentators expect that further availability on other platforms, including on-prem, may be just be weeks away.

Oracle emphasizes that the new release earned the AI moniker owing to its new features clustered around the vector search features.

In a cozy scripted presentation, Oracle founder and CTO Larry Ellison said the vector search features would help specialize foundational large language models with an organization’s own data. “A foundational model is trained, but it doesn’t have your banking records … it doesn’t have your personal email: that’s not public data that these large language models are trained on. But you can specialize these large language models with your personal data [with vector search].”

• Cloud Big Three take lion’s share as market expands 21%
• Intel, Ampere show running LLMs on CPUs isn’t as crazy as it sounds
• Java 17 now developers’ favorite brew, with Eclipse Adoptium proving popular, too
• Oracle Fusion rollout costs 15 times council’s estimates in SAP rip-‘n-replace

When Oracle teased the native VECTOR datatype and “optimized” vector similarity search indexing last year, it won praise from analysts, but the approach is far from unique in the database market. PostgreSQL, MongoDB, Cassandra, and SingleStore all have vector search on their main databases, while vector database vendors such as Pinecone argue that a specialist database will pay off under demanding workloads.

Carl Olofson, research vice president at IDC, said isolated databases can become yet another data resource to manage and keep synchronized. In a statement, he said Oracle Database 23ai aimed to solve the problem by enabling the coordination of vectors with customer’s data using AI Vector Search. “It provides an integrated set of functionality customers need to ground their generative AI, keeping the content current and relevant as they move from projects to production, leveraging all the advances inherent in the database.”

The Oracle Database 19c – the precursor to 23ai on the main upgrade path – will be supported by Oracle under Extended Support until April 2026, giving users around two years to consider their options. Third-party support is available for many older Oracle products. Oracle is offering a free-tier version of 23ai in Oracle Cloud Infrastructure. ®