Oracle says its New Mexico and Wisconsin data-center projects remain on track, with permitting and grid-readiness evolving. Management maintains that neither project will threaten the fiscal 2027 revenue outlook, supported by a diversified, phased capacity pipeline and robust RPO growth; ongoing financing and BYOH arrangements reduce direct capex exposure, though broader, multi-site execution risks remain.
Generated by Dafinchi AI. Source-grounded AI analysis, not investment advice.
What is the current status of New Mexico and Wisconsin data center projects, and could delays pose a risk to the 2027 revenue guidance? As RPO grows, how confident is Oracle in securing capacity online to support future growth?
New Mexico appears operationally on track. Oracle said construction is making “very good progress” and that the data center is “definitely on track.” The remaining notable item is the air-permit process, which Oracle is pursuing with local regulators and the Dona Ana County community. The site is expected to use Bloom fuel cells for on-site power, with management citing low water consumption and low emissions as advantages. 1
Wisconsin is also described as on track. Unlike New Mexico, the project is not planned around on-site generation; Oracle is working with the grid and partners including the Public Service Commission, American Transmission Company, and We Energies to design and deliver the required power infrastructure. Oracle characterized data-center delivery in Wisconsin as “actually very on track and going well,” while acknowledging that the energy-design and delivery plan continues to evolve. 1
These are important projects, but they are not Oracle’s only sources of capacity. Management described New Mexico and Wisconsin as large sites of approximately one gigawatt each, but said Oracle has a broad portfolio of developments across the United States and internationally; neither site contributed to the 850 megawatts delivered in Q1. 2 Oracle also emphasized that capacity at large campuses is delivered in phases over multiple quarters rather than arriving all at once, which reduces the risk that a delay at one site would create a single-quarter capacity shortfall equal to the entire project. 2
Management’s explicit answer is no. Oracle stated that neither New Mexico nor Wisconsin would affect its previously stated fiscal 2027 revenue or earnings guidance. The company had raised its full-year fiscal 2027 revenue outlook to at least $90 billion. 34
That assurance is supported by several factors:
Diversified delivery pipeline: Oracle delivered 850 megawatts of AI capacity in Q1, despite New Mexico and Wisconsin not yet being included in that delivery. 2 It also delivered more than 3,000 GPUs during the quarter, with Q1 delivery nearly three times Q4 delivery and equal to 73% of the prior fiscal year’s total capacity delivered. 5
Phased project ramp: Large campuses are expected to come online over multiple quarters, so a delay at one location should not automatically remove a full gigawatt of capacity from a single reporting period. 2
Risk-adjusted planning: Management said it does not assume every project milestone will be completed exactly on time, and that it maintains backup options and incorporates execution risk into its forecasting. 1
The appropriate interpretation is that Oracle does not currently view New Mexico or Wisconsin as a fiscal 2027 guidance risk, rather than that delays are impossible. New Mexico still has a permitting process underway, and Wisconsin’s grid-delivery plan remains subject to ongoing coordination with utilities and regulators. 1 Oracle also expressly cautioned that its forward-looking statements are subject to risks and uncertainties that could cause actual results to differ materially. 6 Thus, the principal risk would be a broader or simultaneous failure across multiple sites, power arrangements, or alternative capacity sources—not merely a moderate delay at either individual project.
Oracle expressed high confidence, stating that it remains “very excited and very confident” in its ability to meet both current RPO and expected future RPO growth. Management said the company is pursuing multiple avenues to bring capacity online as constraints have shifted from GPUs and semiconductor fabrication to power generation and data-center availability. 4
That confidence is supported by the following evidence:
Strong current utilization and demand: GPU utilization was 97.9% in Q1, and more than $30 billion of additional AI contracts were signed without requiring additional capital from Oracle. 5
Evidence of durable asset economics: All GPU capacity up for renewal in Q1 was either renewed or resold at a 20% premium to prior contracts, and the majority of those GPUs were at least four years old. 5 This suggests that Oracle is seeing continued demand and resale value rather than rapid obsolescence in its deployed AI infrastructure.
Substantial execution capability already demonstrated: Oracle delivered 618 megawatts, representing 75% of the capacity, across six of eight Abilene campus buildings, and delivered 131,000 GPUs there in Q1, 1.9 times the Q4 volume. 5 The next gigawatt-scale campus, Shackleford, was also described as progressing well, with initial NVIDIA Vera system deliveries expected in Q2. 5
RPO growth is not entirely dependent on Oracle-funded capacity: RPO increased by $26 billion from Q4, and management said the vast majority of new contracts used prepayments, bring-your-own-hardware arrangements, or similar structures that do not require incremental cash from Oracle. 7 Oracle clarified that this does not mean the associated infrastructure requires no capital expenditure; rather, suppliers or customers can help fund the hardware through financing arrangements, customer-owned equipment, or upfront payments. 8
The business model is becoming more capital-flexible: Oracle said it has developed supplier relationships, financing arrangements, and bring-your-own-hardware models that separate the capital required for growth from Oracle’s own direct CapEx. Management therefore does not currently view Oracle’s own CapEx capacity as a fundamental limitation on business growth. 9
Oracle’s confidence is credible in light of its current delivery record, diversified project base, customer prepayments, and high utilization. However, the company is operating in a constrained infrastructure environment: management specifically identified power generation and data-center availability as current constraints. 4 New Mexico depends on completion of permitting for its on-site generation, while Wisconsin depends on successful grid and utility coordination. 1
Furthermore, Oracle said that approximately half of its RPO is expected to convert into sales over the next 36 months, while the new RPO added in Q1 is not expected to affect revenue or CapEx until fiscal 2028 or later. 7 This provides some near-term timing flexibility, but it also means future revenue conversion depends on Oracle and its partners continuing to secure power, equipment, sites, and financing at scale.
Disclaimer: The output generated by dafinchi.ai, a Large Language Model (LLM), may contain inaccuracies or "hallucinations." Users should independently verify the accuracy of any mathematical calculations, numerical data, and associated units, as well as the credibility of any sources cited. The developers and providers of dafinchi.ai cannot be held liable for any inaccuracies or decisions made based on the LLM's output.
Oracle’s customer prepayments, customer-provided hardware, and supplier financing can reduce how much cash Oracle must advance as RPO grows. These models change the timing and source of funding, not the need to build capacity. Management expects strong cash generation from projects after ramp-up, but gave no timeline for positive free cash flow and emphasized that infrastructure margins still depend on pricing, costs, and operating efficiency.
Sources used
Research questionHow are Oracle's prepayment, bring-your-own-hardware, and supplier-financing models changing the relationship between RPO growth, CapEx, free cash flow, and infrastructure margins?
Answer outline
Oracle said its global data center buildout is advancing across multiple markets, with 850 megawatts of AI capacity delivered in Q1 and record new capacity brought online. Abilene has made substantial progress, while projects in Shackleford, New Mexico, and Wisconsin continue on differing timelines. Management is planning for phased delivery and execution risks, and said New Mexico and Wisconsin will not change FY2027 guidance.
Sources used
Research questionWhat did management say about Global data center expansion status?
Answer outline
Oracle outlines a path to high steady-state ROIC for OCI infrastructure, supported by contract terms, strong utilization, and megawatt delivery, with near-term gross margin pressure but anticipated rapid improvement as revenue hits full contractual levels.
Sources used
Research questionWhat evidence and operating metrics did Oracle cite to support high steady-state ROIC for OCI infrastructure despite heavy investment, and how did management expect margins to evolve as data center projects reach full contractual revenue levels?
Answer outline
Oracle frames AI infrastructure as a supply-versus-demand contest, noting that demand remains massively higher than available capacity even as many vendors enter the market. The company positions itself on delivering end-to-end, secure, and cost-efficient cloud infrastructure rather than merely selling accelerators, emphasizing operational excellence, full-stack delivery (compute, storage, security, networking, and ongoing management), and high GPU utilization as drivers of competitive advantage and customer satisfaction. The transcript highlights capacity expansion as a lever for renewals and new customer acquisitions: sustained by multi-year relationships, utilization-driven revenue, and the ability to meet rising demand with scalable capacity. It also discusses near-term margin pressure during ramp, followed by rapid uplift as contractual revenue accrues, supported by pricing strategies that pass through uncertain component costs and a long-term ROIC target in the high 20s, with a focus on profitability even as scale increases.
Sources used
Research questionWhat is Oracle's view of competitive dynamics in the AI infrastructure/data-center space, and how will capacity expansion affect customer renewals, new customer acquisitions, and margins?
Answer outline
Oracle attributes Q4 CapEx overspend to timing rather than sustained component-cost pressures, with memory costs acknowledged but not the primary driver. Management explains fixed-price contracts when costs are certain and pass-through mechanisms when costs are uncertain, plus prepaid deals that affect reported CapEx and net cash outlay.
Sources used
Research questionHow did Oracle’s CapEx performance in Q4 (including any component cost pressures like memory) differ from expectations, and what timing vs pass-through mechanisms did management describe for fixed-price versus cost-uncertain long-term contracts?
Answer outline
Oracle's AI infrastructure segment exhibits exceptional growth in Q3 2026, driving significant profitability improvements and exceeding revenue and profit guidance. This expansion is supported by innovative funding and operational efficiencies.
Sources used
Research questionHow is Oracle's AI infrastructure growth expected to impact profitability in 2026?
Answer outline
🚀 Oracle's AI database poised for widespread adoption among large enterprises starting late 2026, driven by cloud expansion, AI integration, and strong security features.🔐🌐
Sources used
Research questionWhat is the expected timeline for widespread adoption of the Oracle AI database among large enterprise customers?
Answer outline
🚀 Dive into Oracle Corp's bold initiatives as they ramp up AI and infrastructure investments to stay ahead of the market! 💡
Sources used
Research questionDiscuss the capex and other investments being made in AI and infrastructure?
Answer outline
Oracle frames the AI data-center market as a full-stack service play anchored in reliability, security, and ongoing operational excellence. Capacity expansion is presented as essential to meet structurally high demand, sustain renewals, unlock new customer acquisitions, and drive margin improvement as utilization reaches full contractual levels. The discussion highlights very high utilization (97.5%), rapid capacity reallocation at renewals, and BYOH/prepaid contracts that support margin stability as the business scales.
Sources used
Research questionWhat is Oracle's view of competitive dynamics in the AI infrastructure/data-center space, and how will capacity expansion affect customer renewals, new customer acquisitions, and margins?
Answer outline
Oracle's Q3 2026 earnings reveal strong SaaS demand fueled by extensive AI integration, underscoring robust backlog growth and multi-industry adoption. Embedded AI agents and new SaaS offerings position Oracle for continued revenue expansion.
Sources used
Research questionWhat do Oracle's demand and backlog signals indicate about AI integration in SaaS applications in 2026?
Answer outline
Oracle plans extensive AI integration across its SaaS applications in 2026, embedding AI agents and launching new AI-powered products to drive growth and automation. The company leverages its AI data platform and multicloud partnerships to enhance enterprise SaaS capabilities and revenue.
Sources used
Research questionWhat are Oracle's plans for AI integration in SaaS applications in 2026?
Answer outline
🚀 Oracle is rapidly scaling its multi-cloud AI data platform by expanding cloud regions, innovating AI data unification, and strengthening security to support diverse, non-Oracle data sources. 🌐🤖
Sources used
Research questionHow does Oracle plan to scale its multi-cloud AI data platform to support growing volumes of non-Oracle data sources and customers?
Answer outline