Pinecone today revealed that an ability to assign workloads to a specific instance of its vector database is now generally available across its managed service.

Jeff Zhu, vice president of product for Pinecone, said the Pinecone Dedicated Read Nodes (DRN) option provides IT teams with more granular control over where workloads run to improve overall performance by enabling a workload to essentially take over an entire node versus running in a multi-tenancy environment.

That approach provides IT teams with more predictable levels of performance for mission-critical workloads, he added.

At the same time, DRN also makes it simpler for IT teams to predict actual costs for a specific workload by isolating it to a specific number of nodes, he added.

It’s not clear how many IT teams are deploying production workloads on a dedicated vector database. Rather than relying on a general purpose database that has been extended to support vectors as an additional data type, Pinecone is making a case for deploying high performance applications on a dedicated vector database designed from the ground up to run those applications at higher levels of scale.

In the case of the DRN service, the actual management of the underlying vector database is also outsourced to Pinecone to reduce the level of expertise required to deploy a type of database that many IT teams lack the skills and expertise to manage at scale.

Each IT team will need to determine to what degree they might require a dedicated vector database. In some instances, for example, an AI application that requires access to a vector database might be built using a general purpose database that has that capability, but deployed on a vector database capable of running that application at scale.

Regardless of approach, the level of investment in databases and other related technologies is expanding in the age of AI. In fact, a recent Futurum Group survey projects the global data intelligence, analytics, and infrastructure (DIAI) market will grow at a 17% compound annual growth rate (CAGR) through 2028 off a base of $541.1 billion in 2026. AI development and operations are specifically forecasted to increase (24%).

Much of that investment is being made to drive a fundamental shift away from manual data engineering workflows in favor of more automated workflows constructed by data engineering teams that are being used to drive AI applications at the level of scale needed to move beyond a simple proof-of-concept.

Ultimately, IT environments are becoming more complex as the volume of data increases alongside the number of formats that data comes in. Many IT teams will inevitably need to revisit their existing data management strategies, which are often mainly focused around structured data rather than the unstructured data that is being used to drive AI applications.

In the meantime, however, IT teams, if they have not already, should become more familiar with the nuances of vector databases that in one form or another are increasingly being more widely distributed across IT environments.