Oracle’s cloud infrastructure business expanded revenue by 31% compared with the prior year, driven largely by demand for high-performance computing workloads and AI use cases. The cloud segment now accounts for 49% of Oracle’s overall revenue mix. Enterprise customers in financial services and life sciences have collectively added more than 200 new OCI instances over the last quarter, reflecting accelerated migration of latency-sensitive applications and data analytics operations to Oracle’s cloud platform.

Oracle reports a backlog of unrecognized revenue totaling $523 billion, bolstered by a landmark $300 billion multi-year agreement with OpenAI to support GPU-accelerated model training and inference. The deal secures long-term commitments for Oracle’s next-generation data centers in the U.S., Europe and Asia, accounting for roughly 60% of the current backlog value. Management expects to recognize approximately 20% of the backlog as revenue over the next three fiscal years as deployment milestones are achieved.

Oracle has announced plans to invest nearly $50 billion in capital expenditures on data centers, GPU clusters and network infrastructure through fiscal 2027. This commitment represents a 70% increase over the prior multi-year cycle and includes expansion of three new high-density cloud regions in India, Brazil and South Korea. The capital program is designed to support projected AI workload growth and reduce time-to-deployment for enterprise clients by up to 40%.

A weather-related power failure at one of Oracle’s primary data centers in Virginia disrupted TikTok’s U.S. operations for several hours. Although emergency generators restored service within two hours, the incident prompted Oracle to fast-track redundancy upgrades and enhanced grid interconnects across eight key regions. The company has allocated $200 million from its CapEx budget to bolster critical power infrastructure and ensure 99.995% uptime SLA compliance for top-tier cloud customers.