Oracle’s multicloud momentum has moved from bold idea to measurable revenue engine, and the numbers Larry Ellison presented at Oracle’s Financial Analysts Meeting make that clear:
multicloud database revenue surged dramatically in the most recent quarter and, according to Ellison, most of that surge came from the Microsoft Azure partnership — while Google Cloud and AWS are only now starting to meaningfully contribute. This shift matters because it changes how enterprises can buy and run mission‑critical databases, accelerates AI adoption on enterprise data, and forces every hyperscaler to answer a new market standard: interoperability without walled gardens.
Background / Overview
Oracle framed the story in precise commercial terms: a reported multicloud database revenue increase in the most recent quarter measured in the
triple‑digit to quadruple‑digit range, driven initially by Oracle Database deployments running inside Azure datacenters under the Oracle Database@Azure model. The company has since extended the same multicloud model — colocated Oracle Database services run and managed by Oracle inside other vendors’ datacenters — to
Google Cloud and
Amazon Web Services (AWS) through Oracle Database@Google Cloud and Oracle Database@AWS.
This “database‑in‑your‑cloud” approach does three important things for enterprise customers:
- It preserves the Oracle database investment and feature set.
- It puts Oracle Database services physically closer to apps and AI services hosted in the hyperscaler of choice.
- It simplifies procurement and operations by letting customers buy Oracle database services via the public cloud vendor they prefer.
Larry Ellison’s public comments at the Financial Analysts Meeting connected the dots between those commercial moves and Oracle’s larger strategic bet: the AI era will massively expand demand for secure, enterprise‑grade inferencing (AI reasoning) on proprietary enterprise data, and Oracle’s database plus multicloud distribution is a unique path to deliver that capability at scale.
Why this multicloud thrust is different — and why it’s working
From “run Oracle on a cloud” to “Oracle runs in the hyperscaler”
Historically, enterprises could run Oracle software on a hyperscaler, but that usually meant a customer‑managed instance on virtual machines or a managed database service provided by the hyperscaler (e.g., RDS for Oracle). The recent wave of announcements represents a different model:
Oracle operates and manages its database services inside the hyperscaler’s own datacenters, delivering parity with Oracle Cloud Infrastructure (OCI) database services and tightly integrating with the hyperscaler’s ecosystem.
The practical results are compelling for enterprises:
- Lower latency for data‑heavy apps when database and app compute are colocated.
- Unified support and single‑pane management for Oracle database services, even when the apps live on a different cloud.
- The ability to buy Oracle database services through the hyperscaler’s procurement channels, leveraging existing cloud contracts and discounts.
Four customer‑driven standards that reshaped the market
Over the last 12–18 months, vendors have adjusted to requirements the enterprise market now demands. Oracle’s multicloud program embodies — and in some cases set — these standards:
- Simplify extraction of value from enterprise data. Enterprises want fewer integration headaches and faster time to insight.
- Expand vendor choice. Customers demand the freedom to pair the best AI, analytics, and platform services from multiple providers without rip‑and‑replace.
- Shorten time to value. Fast deployment, single billing paths, and managed services reduce project risk and accelerate adoption.
- Dismantle walled gardens. Enterprises reject architectures that force a binary choice between cloud vendors; interoperability is now a buying factor.
Oracle’s multicloud approach hits those marks: it lowers the friction of migration, offers choice in where workloads live, shortens path to AI value by enabling low‑latency inferencing against enterprise data, and reduces the lock‑in calculus enterprises once faced.
The mechanics: Azure first, then Google and AWS
Azure: the early mover advantage
Microsoft and Oracle announced a deep partnership that placed Oracle Database services inside Azure datacenters. Because this was the first major implementation of the “Oracle‑operated database in another hyperscaler” model at scale, Azure got the early wave of enterprise orders and integration work. That head start shows up in the near‑term revenue mix — Ellison noted that the vast majority of the quarter’s multicloud growth was driven by Azure.
Why Azure surged early:
- Joint go‑to‑market motions between Microsoft and Oracle accelerated uptake among large enterprise customers already invested in Microsoft services.
- Azure’s enterprise sales channels were able to sell Oracle Database@Azure through existing commercial and contract mechanisms, increasing purchase friction to a minimum.
- Collocation of Oracle Exadata hardware inside Azure datacenters delivered the performance profile that enterprises demand for mission‑critical workloads.
Google Cloud: engineered interconnect and partner programs
Oracle’s work with Google Cloud centered on private interconnects (low‑latency, zero or mitigated egress economics) and rollout of Oracle Database@Google Cloud in target regions. The Google partnership built in features aimed at AI workloads — tight integrations with Google’s analytics and Vertex AI capabilities — which matter for customers whose analytical pipelines and AI models live on Google Cloud.
Google’s strength in AI tooling and analytics makes the partnership attractive for customers who want Oracle’s database as a trusted data layer feeding Vertex AI and BigQuery pipelines.
AWS: a late but meaningful entrant
AWS and Oracle likewise announced Oracle Database@AWS, with Oracle operating Exadata and Autonomous Database services inside AWS datacenters. AWS brings the world’s largest cloud footprint and a massive installed base of enterprise apps, so this cadence — even if it started later than Azure’s — is strategically critical. AWS’s channel muscle and broad set of data services (analytics, S3, Bedrock) make it a natural complement for customers that want Oracle Database paired with the breadth of AWS services.
Oracle’s AI Database and the inferencing thesis
Ellison’s central AI claim is that
AI reasoning (inferencing) will be vastly more ubiquitous than AI training. Training large models is today the province of a handful of AI firms and massive cloud customers; inferencing against enterprise data is a near‑universal requirement for businesses seeking to operationalize generative AI.
Oracle’s product moves are consistent with that thesis:
- The company introduced an AI Database that tightly couples Oracle’s data management features with vector indexes, semantic search, and integrated inferencing pipelines.
- Oracle offers generative AI services inside OCI and ensures those services can access enterprise data without expensive ETL, and without moving raw data to a third‑party LLM provider.
- Multicloud distribution lets enterprises perform inferencing closer to where their application logic and model serving run — reducing latency and enabling consistent governance.
The combined implication: enterprises get managed Oracle database features plus an integrated route into major LLMs and inference stacks, without having to rearchitect their data estate.
Financial framing: can Oracle’s database business hit $20 billion?
Ellison floated the idea — to laughter at a financial meeting — that the multicloud business
might on its own move the Oracle database business toward a $20 billion run‑rate in five years. This is plausible but contingent on several execution factors:
Key variables that will determine the plausibility of the $20B outcome:
- The base database revenue today (the Oracle database franchise still generates high single‑digit to low‑double‑digit billions annually between licenses, support, and cloud database services).
- The sustainability of multicloud growth: one quarter of 1,500% growth is powerful but likely reflects a small initial base; sustaining large percentage gains requires continued expansion into new regions and broad enterprise uptake across Google and AWS on top of Azure.
- AI adoption velocity: the pace at which enterprises adopt inferencing at scale will directly multiply database consumption (vector indexes, embeddings, frequent calls to data for LLMs).
- Pricing and margin mixes: how Oracle prices multicloud database services, whether customers use consumption or subscription models, and the margin pressure of running managed infrastructure inside third‑party datacenters will affect reported database revenues.
- Contracting and channel economics: allowing hyperscalers to sell Oracle database services via their marketplaces increases distribution but may shift margin and requires careful management of reseller and commission models.
In short, the $20B target is achievable under optimistic assumptions — continued multicloud regional expansion, sustained enterprise AI consumption, and Oracle’s success at preserving a favorable pricing and support mix. But the path is not automatic and requires successful execution across product, infrastructure, and commercial fronts.
Strategic strengths behind Oracle’s surge
- Legacy franchise, modernized delivery. Oracle’s installed base of enterprise databases gives it an immense addressable market; the multicloud model turns that legacy asset into a modern cloud subscription cash engine.
- Managed‑service parity. By operating Oracle services inside a hyperscaler datacenter, Oracle avoids the “you get less feature parity on the hyperscaler” complaint that has long inhibited migrations.
- AI‑first product integration. Embedding vector search, cataloging, and inferencing hooks into the database itself lowers the operational barrier to AI use on proprietary data.
- Commercial flexibility. Making Oracle services purchasable through Azure, Google Cloud, and AWS marketplaces reduces procurement friction and shortens sales cycles.
- Vendor neutrality for the customer. Enterprises no longer have to choose between keeping Oracle on‑premise or porting to a hyperscaler’s totally different database environment; they can keep Oracle and still capture the cloud and AI benefits of their hyperscaler of choice.
Material risks and open questions
No strategy is without risk. Oracle’s multicloud play raises a set of operational, commercial, and regulatory questions that enterprises and partners should consider.
1. Margin erosion and channel economics
Running managed Oracle database services inside another vendor’s datacenter implies shared economics. Resale arrangements, marketplace discounts, and hyperscaler commission structures can compress margins compared with Oracle‑operated OCI deployments. Oracle must balance share gain against margin dilution.
2. Sales incentives and channel conflict
The arrangement where hyperscaler sales teams can sell Oracle Database services through their own channels
creates potential channel conflicts — and while that’s great for customers, it requires careful co‑selling governance. Reports that hyperscaler sellers can get commissions on Oracle database sales are credible and explain part of the early success, but also create incentive systems that must be managed to avoid perverse behaviors or over‑promotion.
3. Complexity for IT teams
Multicloud removes vendor lock‑in in one sense but introduces architectural complexity in another. Enterprises will need:
- Clear governance for where data lives and how it flows between clouds.
- Strong identity, access, and encryption practices across cloud boundaries.
- Observability and cost‑management across multiple billing and quota regimes.
4. Latency, data locality and sovereignty
While collocation mitigates latency, cross‑cloud deployments still raise questions for highly regulated workloads that demand strict data residency. Oracle and hyperscalers will need continued regional rollouts and compliance controls to meet sectoral regulations.
5. Reliance on hyperscalers for distribution
Oracle’s model depends on hyperscalers opening their marketplaces and datacenters to Oracle operations — a relationship that is fundamentally cooperative but can be influenced by strategic friction, changing commercial terms, or regulatory pressure.
6. Antitrust and regulatory visibility
A new model where the major hyperscalers resell and host a direct competitor’s managed services creates an unusual competitive dynamic and may draw regulatory scrutiny in markets that are sensitive to dominant platform behavior and resale agreements.
7. The assumption that inferencing is the universal demand driver
Ellison’s claim that “everyone will want to do AI reasoning” is persuasive as a broad business trend, but timing and depth of enterprise inferencing adoption vary widely by sector. Some industries move cautiously for risk, privacy, or cost reasons.
What this means for enterprises and IT decision‑makers
For CIOs and IT procurement teams, Oracle’s multicloud strategy presents actionable choices today:
- Reassess data‑gravity assumptions: enterprises no longer need to rip out Oracle to run modern AI workloads in a hyperscaler.
- Evaluate Oracle Database@<cloud> offerings regionally, including latency tests and cost modelling (consumption vs subscription).
- Revisit procurement clauses: marketplace purchases can simplify buying but need careful contract review, particularly around support SLAs and disaster recovery.
- Strengthen cross‑cloud governance: identity, encryption, data catalogs, and observability are now table stakes.
- Pilot inferencing workloads: run small‑scale inferencing against enterprise data to measure latency, cost per inference, and governance leak points.
How hyperscalers and competitors will react
Oracle’s multicloud push changes competitive optics for hyperscalers and database vendors alike:
- Hyperscalers will need to develop more open resale and marketplace experiences to avoid ceding enterprise database revenue to managed competitors.
- Other database vendors (commercial and open‑source) will emphasize cloud‑native integrations and multi‑cloud data fabrics to maintain competitive parity.
- AI model providers and platform teams will seek tighter integrations with database layers and lower‑latency interconnects, intensifying investments in cross‑cloud networking.
Expect the market to respond with:
- Faster interconnect rollouts and region expansion.
- New partner programs that resell third‑party managed services.
- Competitive pricing moves to make switching or running hybrid setups more attractive.
Bottom line: a customer‑first market, but execution will decide winners
Oracle’s multicloud boom is real and technically coherent: the company preserved core database value while making it available from the hyperscaler of the customer’s choosing. That combination — a trusted database experience + low friction cloud procurement + AI integrations — explains rapid adoption in early markets, especially through Azure.
But the long arc to multi‑billion outcomes depends on execution across several fronts: completing regional multicloud infrastructure, managing channel economics, protecting margins, enabling consistent governance for customers, and scaling AI inferencing use cases in a way that customers are willing to pay for.
For enterprises, the practical benefit is straightforward: more choice, fewer forced migrations, and a simpler path to operationalizing AI on enterprise data. For Oracle and the hyperscalers, this is the start of a new competitive playbook: cooperation where customers demand it, competition where differentiation remains possible.
This is a tectonic shift in the cloud era — not because any single quarter’s headline growth guarantees long‑term dominance, but because the market has formally acknowledged that
interoperability, customer choice, and AI‑ready data platforms are no longer optional. The companies that can operationalize those three elements while preserving economics and governance will define cloud computing’s next era.
Source: Cloud Wars
Oracle is using its multicloud partnerships with Microsoft, Google and AWS to reignite its core database business. It aims to reach $20 billion in revenue within five years by riding the AI inference wave and offering flexible multicloud deployment.