For decades, Oracle was seen as a legacy software stalwart, best known for databases, ERP systems, and sticky licensing contracts. By contrast, Amazon Web Services, Microsoft Azure, and Google Cloud defined hyperscale innovation, while Oracle’s late cloud pivot looked half-hearted.

That changed in 2024. Oracle began rapidly expanding AI-focused infrastructure. Oracle Cloud Infrastructure (OCI) launched competitively priced GPU clusters designed for training and inference of large AI models, quickly winning deals with start-ups and established tech firms.

By mid-2025, analysts described Oracle’s strategy as a “winning formula” in AI compute. Its unusual embrace of multi-cloud partnerships allowed enterprises to run workloads across providers, even competitors. Combined with an aggressive global datacenter buildout, Oracle’s credibility surged.

The Q1 FY2026 release capped this pivot. Legacy software stayed weak, but cloud services revenue rose 28% YoY to $7.2bn, with OCI up 55% to $3.3bn. Most importantly, Oracle’s Remaining Performance Obligation (RPO, or the value of future revenue that is already contracted) exploded by a whopping 359%, recasting Oracle overnight from cloud follower to AI kingmaker.

Wall Street expected a 69% rise in RPO; instead, it got 359%. Oracle added more than $300bn of new contracts in one quarter, well beyond the most bullish forecasts.

Oracle secured four multi-billion-dollar AI cloud contracts, including a blockbuster deal with OpenAI reportedly worth $300bn over five years to support its “Stargate” datacenter project. Elon Musk’s xAI, which trains Grok models on OCI, was another contributor. Partnerships with NVIDIA, Cohere, and even rival hyperscalers (Microsoft, Amazon, Google) broadened the portfolio.

Oracle is also tapping sovereign demand. Governments in Europe, the Middle East, and Asia are pouring billions into domestic AI infrastructure to reduce dependence on U.S. providers. With sovereign cloud offerings tailored for compliance and data residency, Oracle has emerged as a natural partner in this geopolitical race.

The diversity of these contracts, spanning start-ups to superpowers, underscores the breadth of Oracle’s AI opportunity.

Chart 1: Oracle Remaining Performance Obligation surged in fiscal quarter ending August 2025

Source: The Transcript, Oracle reports

The backlog shock triggered one of the most dramatic single-day stock moves in technology history. Oracle was not heavily shorted (short interest just 1.35% of float), but sentiment was cautious given strong year-to-date gains. Investors expected little more than a soft cloud beat, and Q1 numbers were underwhelming: revenue of $14.9bn rose 12% YoY, slightly below consensus, while EPS of $1.47 matched forecasts. Guidance for next quarter was also below expectations and management confirmed that free cash flow would stay negative throughout this fiscal year due to heavy cloud capex.

Yet the RPO figure blew a hole through consensus models. Underweight funds scrambled to buy, sparking a momentum-driven rally. Analysts likened the reaction to Nvidia’s August 2023 earnings, when 171% growth in datacenter revenue sent its stock up 24% overnight.

This time the effect was even broader. Oracle’s surge ignited gains across the AI supply chain. Broadcom rose nearly 10% on optimism for its custom AI chips, Arm Holdings gained a similar amount, Nvidia advanced 4%, and names from Vertiv to Super Micro Computer climbed as well. The market read Oracle’s results not just as a one-off, but as confirmation of a structural AI boom.

Chart 2: Oracle share price (12 months) – 36% surge on 10 September following the Q1 FY26 report

 Source: FactSet

Crucially, Oracle did not present this as a one-off. Management raised FY26 revenue guidance to at least $66bn, with OCI expected to hit $18bn, a 77% YoY increase.

Looking further ahead, Oracle guided OCI revenue to $32bn in FY27, $73bn in 2028, $114bn in 2029, and $144bn in 2030. This implies CAGR above 50%, which is almost unheard of in large-cap tech.

Sceptics warn of overbuilding and balance-sheet strain. But with multi-year contracts already signed, much of this growth is effectively secured. Oracle’s backlog provides visibility unmatched by peers, reinforcing arguments that AI capex is a structural shift akin to the internet buildout two decades ago.

Chart 3: Oracle Cloud Infrastructure Growth – new guidance vs. prior expectations

Source: Bloomberg Intelligence, Oracle reports

Debate this year has centered on sustainability: risks of overcapacity, uncertain monetization, and the drag of higher capex. Execution risk is real: building expensive datacenters at this scale leaves little room for error. Rising interest rates or regulatory headwinds could slow adoption. Valuation worries are also valid after such strong stock moves. But AI firms today are not selling promises, they are booking contracted revenue tied to real adoption. That distinction from dot-com history matters. Oracle’s RPO explosion shows demand is arriving faster and larger than expected.

The broader message: AI is no longer hype. It is a multi-year, multi-sector growth engine. Oracle has planted its flag near the top, but the opportunity spans the ecosystem.

• AI chip design: Nvidia remains the leader, and Oracle’s guidance validates Nvidia’s CEO Jensen Huang’s forecast of $3–4tn in global AI infrastructure spend by 2030. We also note Broadcom’s recent strong numbers and bullish comments on its custom AI chips business, while AMD and Arm remain credible challengers.
• AI chip manufacturing: TSMC is the clear winner, with unmatched scale and technology.
• Datacentre hardware: multiple companies are poised to benefit, including names such as Arista (high-speed interconnects), Vertiv (cooling), Amphenol (cabling), and server makers like Super Micro Computer and Dell
• Large language models: privately owned start-ups dominate this space, but among listed stocks, we note that Alphabet is no longer viewed as a laggard in AI models / chatbots, thanks to the significant progress it has made with its Gemini product.
• Cloud hyperscalers: Microsoft, and Alphabet have seen a reacceleration in their cloud business, Amazon trends were stable but near-term comments were bullish too. Chinese hyperscaler Alibaba has also printed strong surge cloud revenue. Moreover, there are new entrants in this space, from Oracle, to more recent ones such as Coreweave or Nebius.
• AI software / Agentic AI:  AI monetization will come in phases: Nvidia is benefitting from the datacenter buildout stage, Oracle, Google and Microsoft are benefitting from renting AI infrastructure. We believe software firms stand to benefit as AI adoption widens, and today Salesforce and ServiceNow appear to have the most advanced agentic AI products in this space.