Microsoft (MSFT.US) and Databricks extend strategic cooperation to the 2030s to jointly promote large-scale implementation of enterprise AI

The Zhitong Finance App learned that on July 23, Microsoft (MSFT.US) and data and artificial intelligence company Databricks announced the expansion of their strategic partnership for more than 10 years, extending the cooperation period to the 2030s, with the aim of helping enterprises build and deploy enterprise-level AI applications based on their own business data.

According to the agreement, Databricks will further deepen the use of the Microsoft Azure cloud platform. Its core business operations and unified lake warehouse architecture will operate based on Azure Databricks, and adopt Microsoft's next-generation Azure Cobalt infrastructure based on the Arm architecture to improve performance and energy efficiency. Azure Databricks is a solution jointly developed by both parties to unify the integration of data, analytics and AI on the Azure cloud platform.

At the infrastructure level, Databricks will expand the use of Microsoft's self-developed Arm-based chip Azure Cobalt. The company currently runs workloads on the Cobalt 100 processor and plans to further adopt the Cobalt 200 — a chip that Microsoft says can provide up to 50% performance improvement over previous generations and enables memory encryption by default. This move shows that in a context where demand for AI computing power continues to rise, the two sides are addressing the performance challenges of data-intensive and AI proxy workloads through customized chip optimization.

The two companies said that Microsoft will also continue to integrate DataBricks' data and AI platform capabilities into all of its products, and directly embed functions including AI collaborator Genie into customer workflows.

Ali Ghodsi, co-founder and CEO of Databricks, said, “By deeply integrating Databricks Genie and Unity AI Gateway into the Microsoft product system, we are helping companies unify data assets and root AI on business knowledge. This enables customers to fully unlock the value of agents and models while effectively controlling costs and ensuring governance compliance.”