Oracle posted strong fiscal Q2 2026 results as cloud momentum accelerated and large AI-related commitments drove Remaining Performance Obligations (RPO) to a record $523 billion, up 438% year-over-year. Revenue climbed 14% to $16.1 billion and cloud services rose 34%, but the company also reported a sharp increase in capital expenditures tied to its global AI data center build-out—$12 billion in Q2 alone—and signaled that full-year FY26 CapEx will be roughly $15 billion higher than previously forecast. While operational metrics were solid, Oracle held its FY26 revenue outlook steady and pushed more upside into FY27, trends that contributed to a cautious market reaction following the earnings release. GAAP EPS rose 91% to $2.10, boosted by a $2.7 billion gain from Oracle’s divestiture of its Ampere stake.
Management signaled a strategic pivot toward full chip neutrality, moving away from designing proprietary processors and emphasizing partnerships with NVIDIA and other CPU and GPU suppliers. Executives also highlighted rapid expansion of Oracle’s global data center footprint, with 211 live and planned regions and 72 multicloud data centers being embedded inside AWS, Google Cloud, and Microsoft Azure. Oracle reported multicloud database revenue jumping 817%—its fastest-growing segment—driven by customers running Oracle databases across multiple cloud platforms.
Oracle also spotlighted AI as a catalyst across its database, analytics, and applications stack. The company sees its greatest opportunity in embedding AI into core workflows such as financial services, logistics, and healthcare. With all top five AI models hosted on Oracle Cloud, executives emphasized advantages in performance, automation, and operational efficiency.
• Q2 revenue reached $16.1B, up 14% in USD.
• Cloud revenue hit $8.0B, up 34%.
• Cloud Infrastructure (IaaS) revenue grew to $4.1B, up 68%.
• Cloud Applications (SaaS) revenue reached $3.9B, up 11%.
• Fusion Cloud ERP revenue rose to $1.1B, up 18%.
• NetSuite ERP revenue reached $1.0B, up 13%.
• RPO surged to $523B, up 438% year-over-year.
• GAAP EPS climbed 91% to $2.10; Non-GAAP EPS rose 54% to $2.26.
• Operating cash flow for the past twelve months totaled $22.3B, up 10%.
• Oracle now counts 211 active and planned cloud regions and is more than halfway through deploying 72 multicloud data centers inside AWS, Google Cloud, and Azure.
• Multicloud database revenue grew 817% year-over-year.
• Oracle’s exit from Ampere reflects a long-term move to chip neutrality.
“Oracle sold Ampere because we no longer think it is strategic for us to continue designing, manufacturing and using our own chips in our cloud datacenters,” said Oracle Chairman and CTO, Larry Ellison. “We are now committed to a policy of chip neutrality where we work closely with all our CPU and GPU suppliers. Of course, we will continue to buy the latest GPUs from NVIDIA, but we need to be prepared and able to deploy whatever chips our customers want to buy. There are going to be a lot of changes in AI technology over the next few years and we must remain agile in response to those changes.”
“Oracle is very good at building and running high-performance and cost-efficient cloud datacenters,” said Oracle CEO, Clay Magouyrk. “For years Oracle has been investing in AI and building autonomous cloud software. Oracle’s Autonomous Database and Autonomous Linux have been key to reducing human labor and human error in our datacenters. Because our datacenters are highly automated, we can build and run more of them. Oracle has over 211 live and planned regions worldwide—more than any of our cloud competitors. We are more than halfway through building 72 Oracle Multicloud datacenters to be embedded throughout the Amazon, Google and Microsoft clouds. We are committed to Cloud Neutrality because we believe that our customers should be able to run their Oracle databases in any cloud they choose. That strategy is definitely paying off. Our Multicloud database business is our fastest growing business—up 817% in Q2.”
Oracle Q2 FY26 Investor Call Highlights
- OCI now runs 147 live customer-facing regions with 64 additional regions planned, underscoring the pace of global AI data center expansion.
- In the quarter, Oracle handed over close to 400 MW of data center capacity to customers and delivered 50% more GPU capacity than in Q1.
- Oracle’s Abilene, Texas AI supercluster has received more than 96,000 NVIDIA Grace Blackwell GB200 GPUs, and the company also began delivering AMD MI355 capacity to customers, adding further diversity to its AI silicon stack.
- Q2 CapEx reached $12 billion, driving free cash flow to negative $10 billion, with the vast majority of spend tied to revenue-generating equipment (GPUs, servers, networking, optics) rather than land, buildings, or power.
- Oracle reiterated that it incurs no cash expenses for large data center facilities and utilities until those are fully delivered and fit for purpose, then purchases equipment late in the build cycle to quickly convert CapEx into revenue.
- Management highlighted flexible funding models that reduce external borrowing needs, including customer “bring-your-own-chip” deployments and vendor arrangements where chips are rented rather than sold.
- Based on current plans and these alternative funding mechanisms, Oracle expects to raise “less, if not substantially less” capital than the ~$100 billion AI build-out figure cited in some analyst reports, while preserving its investment-grade debt rating.
- Oracle continues to target 30–40% gross margins over the life of AI workload contracts on OCI and described the low-margin “ramp” phase for individual AI data centers as typically lasting only a couple of months.
- The company framed AI margin expansion as primarily a utilization and mix problem: as more of the global AI data center fleet reaches steady-state usage, aggregate AI margins should move toward the 30–40% band.
- OCI’s AI capacity is highly fungible: if capacity is reclaimed from a customer, Oracle can securely wipe and reprovision that hardware to another tenant in hours.
- Large AI customers typically ramp new capacity on OCI in two to three days, and Oracle described such reallocation and resizing of GPU clusters as an everyday operational pattern across its 700+ AI customers.
- Dedicated Region and Alloy consumption grew 69% year-over-year, with 39 dedicated regions live and 25 more planned, and new Alloy regions launched by NTT Data and SoftBank, reflecting strong demand for “cloud in your data center” and partner-operated cloud models.
- Marketplace consumption increased 89% year-over-year, driven by partners such as Palo Alto Networks (SASE and Prisma Access on OCI), Cybereason, and Newfold Digital, whose SaaS offerings both expand Oracle’s ecosystem and directly drive OCI infrastructure usage.
- Oracle reported more than 400 AI features now live in its Fusion applications, reinforcing its “baked-in AI” positioning across ERP, HCM, SCM, and CX.
- In healthcare, 274 customers are live in production on Oracle’s clinical AI agent, with go-lives measured in weeks and often implemented directly by customers without Oracle’s help.
- Oracle’s new AI-based ambulatory EHR is generally available and has received U.S. regulatory approval, with management signaling a material acceleration in healthcare bookings and revenue in Q3.
- Oracle completed a major sales reorganization that merges Fusion and industry cloud sales teams into a unified “One Oracle” go-to-market motion, enabling larger multi-pillar deals that blend applications, database, and AI infrastructure.
- Cloud applications deferred revenue grew 14%, outpacing the 11% in-quarter cloud apps revenue growth, supporting management’s view that application growth will further accelerate as AI and One Oracle motions scale.
- At peak holiday loads, Uber surpassed three million cores on OCI to handle record traffic, while Temu scaled to nearly one million cores for Black Friday and Cyber Monday campaigns.
- Oracle supported thousands of retail and consumer customers through their largest seasonal peaks, positioning OCI as a platform for high-traffic, AI-enhanced digital commerce.
- TIM Brasil expanded its five-year OCI deal to deepen AI-driven customer experience: 24 AI projects are in motion (seven in production, six more launching soon), with early results including 18% faster issue resolution, 16% higher customer satisfaction, 30% faster service times, and 15% fewer network failure interventions using AI agents and predictive analytics on OCI.