On September 10, Oracle released its fiscal first-quarter results, posting record revenue of $19.3 billion, a 30% year-over-year increase. The cloud infrastructure segment generated $7.4 billion, an impressive 121% jump that surpassed the $7.19 billion consensus estimate. Following the announcement, Oracle's stock climbed nearly 8% in after-hours trading before settling back to a 4% gain. During the subsequent analyst call, CFO Hilary Maxson characterized the quarter with a single word: "acceleration."

Addressing external worries regarding delays at data centers in New Mexico and Wisconsin, Oracle's management dismissed the concerns, asserting that "the two sites do not impact FY27 results" and emphasizing that "not all eggs are in one basket."

Compute Delivery: 850 Megawatts in a Single Quarter, Nearly Triple the Previous Period

When questioned about data center progress during the call, CEO Clay Magouyrk provided specific figures. In Q1, Oracle delivered 850 megawatts of AI compute capacity to customers, encompassing more than 300,000 GPUs. "Q1 delivery was nearly three times the total delivered in Q4, equivalent to 73% of the entire prior fiscal year's output," Clay noted.

The Abilene hyperscale campus alone saw 131,000 GPUs delivered in Q1, 1.9 times the Q4 volume. Six of the campus's eight buildings have now been delivered, representing 75% of its total capacity. Clay specifically highlighted that OpenAI's recently released GPT-6 Astra was trained at Oracle's Abilene facility.

GPU Renewals: 20% Price Hike Even for Hardware Over Four Years Old

Magouyrk stated, "All GPUs up for renewal in Q1 were renewed or resold at a 20% premium over their original contracts. Most of these GPUs had been in service for four years or longer." He also disclosed that Q1 GPU utilization reached 97.9%.

When pressed by an analyst about the useful life of hardware assets, Magouyrk responded, "I've spent nearly my entire career in the infrastructure industry, with 12 years at Oracle. One pattern has never changed: demand for server-side compute and data centers only increases. The AI use case follows this trend, perhaps even more intensely. Achieving a 20% premium on renewals is a very positive signal for demand sustainability, business growth, and profitability."

RPO: Approximately Half to Convert to Revenue Within 36 Months

Maxson disclosed during the call that the Remaining Performance Obligation (RPO) grew by $26 billion sequentially in Q1 and provided a clear conversion timeline for the first time: "We now expect approximately half of our RPO to convert to sales revenue within the next 36 months."

Additionally, new AI contracts signed in Q1 exceeded $30 billion. Magouyrk emphasized that these contracts "do not require additional capital from Oracle," as payment structures such as prepayments and customer-supplied hardware have partially shifted capital expenditure burdens. Maxson further elaborated on these new obligations: "This new RPO won't impact our capital expenditures, and the revenue won't be recognized until fiscal 2028 or later."

Addressing Data Center Construction Concerns Head-On

During the Q&A, analysts directly questioned the potential delays at the New Mexico and Wisconsin data centers. Magouyrk clarified that neither site contributed to the Q1 850MW delivery figure and that they "will not impact our previously stated FY27 revenue or earnings guidance."

He added, "Our data center portfolio under construction is extensive and diversified across the United States and globally. We're not putting all our eggs in one basket." He also explained that large sites are phased in over time rather than going live all at once, so a delay at any single site won't create a cliff-like shortfall in a given quarter.

Regarding New Mexico, he noted construction is progressing well, with the company advancing through the air permit application process. The plan is to use Bloom fuel cells for on-site power generation, which offer "extremely low emissions and very low water consumption." The Wisconsin facility is also "moving very smoothly," with coordination underway with local electricity regulators for grid-supplied power.

Applications Business: AI as an Accelerator, Not a Replacement

Applications CEO Mike Sicilia laid out the growth logic for the SaaS business during the call. "AI is an accelerator, not a replacement for suite applications," he said. Q1 SaaS grew 10% overall, with Fusion up 14% and industry-specific applications growing over 20%.

He also revealed plans to launch an AI-assisted implementation tool at the AI World conference in October, which aims to "compress SaaS deployment cycles from years to months, and from months to weeks." He cited examples where complex healthcare application rollouts were reduced from "high double-digit months" to "single-digit months," and some NetSuite customers went from "double-digit months" to "single-digit weeks."

Raised Full-Year Guidance; Q2 Cloud Growth Expected to Exceed 65%

Maxson announced that Oracle has raised its full-year revenue guidance to at least $90 billion, representing roughly 34% year-over-year growth, with non-GAAP EPS guidance set at $8.10. For Q2, the company expects total revenue growth between 30% and 34% and cloud revenue growth between 65% and 71%.

Capital expenditures for the full year are projected between $90 billion and $95 billion, with net cash capex not exceeding $70 billion. Maxson stated, "Every one of these projects is, in essence, a strong free cash flow project. Once they reach the ramp phase, they can quickly achieve nearly 100% free cash flow conversion of after-tax EBITDA."

Here is the full transcript of the call:

Oracle Fiscal Q1 2027 Earnings Conference Call

Date: September 10, 2026, Company: Oracle, Event: Fiscal Q1 2027 Earnings Call

Opening Remarks

Moderator: Hello and welcome to Oracle's fiscal first quarter 2027 earnings call. This call is being recorded. Please limit yourselves to one question each. I will now turn the call over to Ken Bond, Head of Investor Relations.

Ken Bond: Thank you, Miriam. Good afternoon, everyone, and welcome to Oracle's Q1 FY27 earnings call. Joining me today are CEO Mike Sicilia, CEO Clay Magouyrk, and CFO Hilary Maxson. The press release—including financial results, data tables, supplemental financial metrics, and guidance—is now available on our Investor Relations website. The slides used in today's call and GAAP to non-GAAP reconciliations are also available there.

Please note that today's discussion will contain forward-looking statements, and we will address several important factors related to our business. These statements are subject to risks and uncertainties, and actual results could differ materially. We encourage you to review our latest filings, including our 10-K, 10-Q, and related amendments. We undertake no obligation to update these statements. We will begin with prepared remarks before moving to Q&A. I'll now turn the call over to Hilary.

Financial Highlights

Hilary Maxson: Thank you, Ken. Hello, everyone. As Ken mentioned, you can follow along with our slides and webcast on our website. If I had to use one word to describe this quarter, it would be "acceleration"—we've accelerated execution across the entire company, translating into strong revenue and profit results. Q1 set another record, driven by both cloud infrastructure and cloud applications.

Total revenue reached a record $19.3 billion, up 30% in USD year-over-year. This is the first time Q1 revenue has grown sequentially, signaling the progress of our large-scale infrastructure buildout. Historically, a record Q4 was followed by a softer Q1, but that pattern no longer applies as we accelerate across the entire stack—from infrastructure to database to software.

Cloud infrastructure revenue was $7.4 billion, up 121% year-over-year, reflecting strong execution in bringing record megawatts online, driven by sustained demand for compute and database services. Cloud application revenue grew 10% year-over-year, with Fusion and industry applications growing well above that pace—Mike and Clay will elaborate shortly.

Non-GAAP operating income was $8.2 billion, up 31% year-over-year, driven by strong revenue growth and operating leverage. Gross margin declined as expected, largely due to the acceleration of data center construction and infrastructure revenue growth. However, this pressure was offset this quarter by lower operating costs and robust operating leverage from simplification initiatives. Overall, non-GAAP operating margin was roughly flat at approximately 42%.

Non-GAAP EPS grew 30% year-over-year to $1.92. Finally, RPO increased by $26 billion sequentially. This has two implications: First, we continued to expand RPO to support future revenue. The vast majority of new contracts use prepayments, customer-supplied hardware, or similar mechanisms, requiring no incremental capital from Oracle. This RPO won't impact capex or revenue until FY28 or later. Second, we began to see strong RPO conversion into revenue this quarter, driving the impressive cloud infrastructure performance. We've added slides illustrating the inflection point in RPO converting to revenue and operating profit.

Specifically, cloud infrastructure revenue grew 121% in Q1, up from 93% in Q4, and we expect this acceleration to continue through FY27 as more RPO converts. We currently expect about half of our RPO to convert to revenue within 36 months. On a trailing twelve-month basis, total company revenue growth accelerated by 5 points in Q1 versus Q4. On the same basis, operating profit growth accelerated from 16% in Q4 to 21% in Q1.

Turning to the balance sheet and cash flow: Operating cash flow was a record $23 billion, reflecting strong execution and customer prepayments. Capital expenditures were $28 billion, resulting in negative free cash flow of $5 billion. Net cash capex, excluding prepayments, was $18 billion. Note that annual capex will not be evenly distributed. We continue to expect full-year capex between $90 billion and $95 billion, with net cash capex not exceeding $70 billion. Finally, we are pleased to report that the previously disclosed $20 billion at-the-market equity offering was completed in Q1. I'll now turn the call over to Mike and then Clay.

Cloud Applications Overview

Mike Sicilia: Thanks, Hilary. Let me add more color on the applications business. We continue to see strong customer adoption of our application suites. They are investing in trusted, complete solutions that seamlessly integrate AI agents with applications to run their businesses. AI is an accelerator for application suites, not a replacement.

For decades, we've built deep expertise in business process operations across industries, regions, and organization sizes, which gives us unique insight into helping customers succeed. Even before AI, application suites proved their effectiveness. End-to-end automated and standardized processes improved margins far better than custom, point solutions. But this required employees to strictly follow system-defined workflows, which was challenging to maintain consistently across functions, teams, and regions.

AI changes this. AI agents can execute tasks based on existing workflows and business rules, with employees supervising agents, handling exceptions, and applying judgment at critical points. By combining embedded AI with decades of complex business rules, compliance requirements, security models, data models, and customizations, we help customers unlock AI value while keeping data secure and operational guardrails intact. This makes it easier than ever for organizations to harness the full power of their application suites. We are confident in the potential for this new paradigm to deliver significant ROI quickly.

At our AI World conference in October, we will launch a new agentic AI accelerator that redefines how customers deploy Oracle applications—faster, simpler, and at significantly lower cost. AI agents will work alongside Oracle and customer teams to automate implementation at unprecedented scale, compressing SaaS deployment cycles from years to months and months to weeks. This is why I'm confident our applications growth will continue to rise.

Q1 results: We saw strong performance across our SaaS portfolio. SaaS grew 10%, Fusion grew 14%. Oracle Health continues to accelerate, and industry applications grew over 20% year-over-year in Q1. As mentioned last quarter, NetSuite experienced slower decision cycles last fiscal year, but we've launched exciting new products that I'll detail shortly.

Customer highlights from the quarter: Uber Technologies, Stanford University, and MUFG Bank have all gone live and/or expanded on Fusion. PYE Barker Fire Safety selected Oracle's complete application suite. Johnson Controls, Saudi National Bank, GuideWell Mutual Holding, and Petronas all added Fusion agentic applications.

AI usage metrics: Customers used our embedded AI capabilities over 150 million times in Q1, up 42% sequentially. AI agents executed over 3.5 million times in production, nearly doubling sequentially. Over 2,300 AI agents are in production, up 90% sequentially. In Fusion alone, AI usage consumed 900 billion tokens.

On NetSuite: We announced general availability of NetSuite Next, a new AI-driven product delivering a simpler, more powerful agentic experience. Additionally, NetSuite AI Connectors have become one of the fastest-adopted features in NetSuite's history, connecting customer data to leading AI assistants like ChatGPT and Claude, with over 10,000 customers now using it. Every Man Jack estimates it saves them approximately $350,000 and nearly 5,000 hours annually.

On Oracle Health: This month, at our Oracle Health & Life Sciences customer event, we will unveil the new Magenta care management system. It's more than an EHR—it connects clinical research and care, reducing administrative burden and using AI as the user interface. Helping improve healthcare quality and allowing clinicians to focus on patients rather than systems is one of our most meaningful AI missions.

Cloud Infrastructure Overview

Clay Magouyrk: Thanks, Mike. OCI continues to grow rapidly to meet customer capacity demands.

On capacity delivery: Since the end of Q4, we delivered 850 megawatts of AI compute, including over 300,000 GPUs. Q1 delivery was nearly three times Q4's total and equivalent to 73% of last fiscal year's total. This is the result of years of investment across data center design, supply chain, manufacturing, and operations.

Demand remains exceptionally strong: New AI contracts in Q1 exceeded $30 billion without requiring incremental capital from Oracle. Our ability to operate large multi-tenant clusters is a significant advantage, with GPU utilization at 97.9% in Q1. GPU lifespan and value continue to exceed expectations—all GPUs up for renewal in Q1 were renewed or resold at a 20% premium, with most being four years or older.

On the Abilene campus: We delivered 131,000 GPUs in Q1, 1.9 times Q4. Six of the eight buildings are complete, representing 618MW or 75% of capacity. Customer acceptance has been compressed to just 24 hours. OpenAI's GPT-6 Astra was trained at our Abilene facility.

On Shackleford: Our next gigawatt-scale campus is on track. NVIDIA Vera Rubin systems are performing better than expected in quality, yield, and performance. We will deliver the first Vera Rubin systems to customers in Q2.

On database cloud: Multicloud database revenue grew 353% year-over-year, with customer count up 180%. We completed planned Azure and AWS regional expansions, now covering 70 multicloud database regions and 119 availability zones. Oracle Interconnect for AWS is now generally available, providing private, high-speed connectivity to all major hyperscalers with no data transfer fees.

On AI ecosystem innovation: We expanded our partnership with OpenAI, offering OpenAI API access, enterprise ChatGPT, and Codex through the Oracle Cloud Marketplace, including GPT-6 Astra. Gemini models are being integrated into Oracle enterprise applications. We released new Grok reasoning, multimodal, and text-to-speech models. Our open-source catalog continues to expand. We announced a multi-year partnership with Quentinium to deliver its Helios quantum computer on OCI.

On developer tools: Oracle APEX 26.1 now runs over two million active applications. APEX Lang is a new technology representing APEX applications as structured, human-readable definitions. The AI Data Platform now integrates with Codex and Cloud Code, allowing developers to leverage its capabilities directly in their preferred environments.

Overall, Q1 demonstrated how our capabilities reinforce each other. We're delivering data centers and GPU capacity at a pace that seemed impossible a year ago. Clients continue to sign new contracts, renew at higher prices, and run clusters near full capacity. This combination keeps demand growing and reinforces our confidence in the long-term value of what we're building.

Guidance

Hilary Maxson: Thanks, Clay. Now for guidance for Q2 FY27 and the full year.

Q2 guidance: Total revenue growth between 30% and 34% in USD. Cloud revenue growth between 65% and 71% in USD. Non-GAAP EPS between $1.85 and $1.93, up 21% to 25% year-over-year, excluding the Ampere-related gain from last year's Q2.

Full-year guidance: Given Q1's strong execution and accelerating momentum, we are raising our full-year guidance. Total revenue of at least $90 billion, up approximately 34% year-over-year. Full-year non-GAAP EPS of $8.10.

Finally, please mark your calendars for our Investor Day in October. Now, back to Ken for Q&A.

Q&A Session

Moderator: Thank you. Please prepare for questions.

Q1 (Brad Zelnick, Deutsche Bank): Ken, congratulations on your retirement. My question: I fully believe Oracle will remain one of the few leaders in AI infrastructure. You've said FY27 and FY28 are peak capex years, but other players seem to be spending endlessly. Should we consider the possibility of higher peaks beyond FY28? What factors guide your investment decisions? Also, when might we see positive free cash flow return?

Clay Magouyrk: Thanks, Brad. Over the past few quarters, we've discussed innovative ways to finance the business. One way is directly using Oracle's capital, but we've also invested heavily in supplier and partner relationships and created new models like customer-supplied hardware. I think we need to decouple "Oracle's direct capex" from business growth potential. The capital needed to drive growth doesn't all have to come from Oracle. It's not a bottleneck; it's an evolution of our business model as we scale with AI.

Hilary Maxson: Regarding free cash flow, we haven't provided specific timing and won't today. But every project we're pursuing is inherently strong from a free cash flow perspective. Once they ramp, they can quickly achieve about 100% free cash flow conversion of after-tax EBITDA. We'll provide more clarity as we expand in coming quarters. The only variable, as you noted, is further growth capex.

Q2 (Siti Panigrahi, Mizuho): Ken, congratulations. There's been speculation about delays at the New Mexico and Wisconsin data centers, yet you delivered on Q1 targets and raised guidance. Can you provide updates on these projects? Is there timing risk that could affect FY27 revenue? Also, how confident are you in securing capacity for growing RPO?

Clay Magouyrk: Thanks, Siti. New Mexico and Wisconsin are important large sites, each around 1 gigawatt. But we just delivered 850MW in Q1, none of which came from Shackleford, New Mexico, Wisconsin, or Michigan. This shows we have a large, diversified portfolio across the US and globally. These two sites get a lot of attention, but we're not putting all our eggs in one basket. Large sites come online in phases over multiple quarters, so a delay at one site won't cause a cliff in any single quarter.

Also, a plan that assumes everything goes perfectly is what we'd call a "bad plan." We understand the complexities of large-scale infrastructure. Regarding New Mexico: Construction is proceeding well. We're advancing the air permit process and plan to use Bloom fuel cells for on-site generation—among the most environmentally friendly options with very low water usage and emissions. Regarding Wisconsin: We're working with local partners on grid-supplied power. Construction is actually progressing very smoothly. Neither site impacts our FY27 revenue or earnings guidance.

We're very confident in our ability to meet current and future RPO. The constraints have evolved—from GPU supply to power to data center construction—but demand is strong enough that we're pursuing every avenue to bring capacity online.

Q3 (Raimo Lenschow, Barclays): Ken, happy retirement. Clay, can you talk about pricing trends for components and contracts? Does last year's gross margin framework still hold?

Clay Magouyrk: In a supply-constrained market, prices typically only go up. Our costs have risen, so we've raised prices to ensure appropriate returns across all related segments. We don't expect a negative impact on gross margin, and prior guidance remains intact. On hardware lifespan, I've seen a consistent pattern over nearly three decades: demand for server-side compute and data centers only grows. AI follows this trend, perhaps more intensely. Renewing at a 20% premium is strong evidence of sustained demand and profitability.

Q4 (Mark Mordler, Bernstein): Ken, we'll miss you. Can you detail what's driving RPO growth that doesn't require incremental Oracle capital—prepayments and customer-supplied hardware? Are participants mainly AI labs and semiconductor companies, or sovereign cloud customers? Also, can you update us on sovereign cloud progress?

Clay Magouyrk: Let me clarify: I didn't say these projects don't require capital. I said they don't require incremental capital from Oracle. We achieve this through supplier financing arrangements, where we pay for equipment as customers pay us; customer-supplied hardware, where customers buy hardware but leverage our operations expertise; and customer prepayments, where customers fund upfront costs. The customer base is broad—from startups to investment-grade enterprises.

On sovereign cloud: Our Alloy business is performing well with strong partners in Japan, the Middle East, and elsewhere, serving both commercial and government customers. Many sovereign cloud customers are also using our GPU capacity for sovereign workloads. We also have a large, fast-growing general-purpose cloud business that requires less capital than large AI clusters.

Q5 (John DiFucci, Guggenheim Securities): Ken, you've been a true friend. Hilary, gross margin declined this quarter due to capacity coming online before revenue. Operating margin was roughly flat. How should we think about the interplay between gross and operating margin?

Hilary Maxson: Gross margin is an important health indicator—it reflects pricing discipline and cost control. But operating margin is ultimately the metric I focus on for value creation. The current gross margin dynamics are driven by data center acceleration and the mix shift—software has higher gross margins but higher opex, while infrastructure has lower gross margins but significantly lower opex. Operating margin is the better measure of long-term value creation.

John DiFucci: Will gross margin continue to decline or stabilize for the rest of the year?

Hilary Maxson: As I mentioned in Q4, gross margin will decline in a stepwise fashion this year. Our EPS guidance reflects our expectations for operating margin. As the expansion phase concludes, we'd reasonably expect gross margin to stabilize, though we're not providing specific guidance today. More details at Investor Day in October.

Q6 (Brent Thill, Jefferies): Mike, good to see SaaS at double-digit growth. Can you discuss the logic for sustaining this?

Mike Sicilia: Certainly. First, we have highly differentiated end-to-end industry suites across healthcare, retail, telecom, construction, and more. Second, AI is embedded as part of regular application updates, providing a seamless way to achieve AI ROI. Third, Fusion Agentic AI Studio allows customers to build their own AI agents within the same platform. Fourth, AI-assisted deployment is compressing implementation timelines dramatically—healthcare from double-digit months to single-digit months, and some NetSuite customers from double-digit months to single-digit weeks. This accelerates revenue recognition. Finally, our AI Data Platform automates enterprise ontology creation. Our SaaS business is also a great lead generator for IaaS—SaaS customers are also buying OCI. We see SaaS as part of our total solution, not a standalone business.

Q7 (Kirk Materne, Evercore ISI): Ken, I'll congratulate you in person at Investor Day. Mike, can you discuss the AI Data Platform business model? Does it drive incremental database and OCI consumption, or is it standalone software revenue?

Mike Sicilia: The answer is "all of the above." It runs outside of Oracle-only environments—it's not exclusive to Oracle Database or applications. It works with any data source, and we're automating ontology construction for hundreds of data sources. Oracle Database remains the world's most trusted platform for mission-critical data, and this will feed our growing multicloud database business.

Kirk Materne: Do you deploy resident engineers to help customers unlock value?

Mike Sicilia: Yes, we already deploy resident engineers for both AI Data Platform and Fusion Agentic Studio, which share a single control plane on OCI. We see it as an integrated platform. The results—measured in weeks in highly regulated industries—were unimaginable even a year ago.

Closing Remarks

Clay Magouyrk: We expect to report Q2 FY27 results on December 14, 2026. A replay of this call will be available on our IR website for 24 hours. Thank you for joining.

Moderator: Thank you for participating. This concludes today's call. Please disconnect.

Disclaimer: This transcript may contain inaccuracies and is provided "as is" without warranties.