Oracle spent years playing catch-up in cloud infrastructure. On August 11, 2026, it tried a different move: skip the queue entirely by betting on hardware nobody else in the hyperscaler tier has locked down. Oracle and quantum computing company Quantinuum announced a multi-year partnership that will put Quantinuum’s Helios trapped-ion quantum computer inside a U.S.-based Oracle AI data center, accessible through a forthcoming Oracle Cloud Infrastructure (OCI) quantum service. The deal lands in the same month Synergy Research Group clocked global cloud infrastructure spending at roughly $143 billion for the second quarter of 2026, up about 43% year over year, the fastest growth rate the market has posted in eight years.
Those two facts sit next to each other for a reason. Oracle Cloud Infrastructure quantum computing plans arrive at a moment when the entire cloud market is expanding fast enough that even a distant fourth-place provider can carve out a defensible niche. This is a news analysis of what Oracle actually announced, how it stacks up against AWS Braket, Azure Quantum, and Google’s Quantum AI program, and why the timing lines up with one of the strongest cloud-spending quarters in nearly a decade.
Oracle’s Quantum Bet: What Was Announced on August 11
Oracle and Quantinuum unveiled the partnership simultaneously from Broomfield, Colorado, and Austin, Texas, describing it in an official Oracle announcement as a “multi-year strategic partnership to bring quantum computing to Oracle Cloud Infrastructure (OCI).” The core of the deal: OCI customers will get direct access to Quantinuum’s Helios quantum computer, wired into Oracle’s existing high-performance computing and GPU infrastructure so quantum jobs can run alongside classical simulation and AI workloads instead of sitting in a separate silo.
Quantinuum’s own account of the announcement, posted the same day, framed the goal as accelerating “the commercial adoption of hybrid quantum-AI computing.” Oracle has not published pricing or a firm launch date for the OCI quantum service. The company says it plans to preview the service “in the coming months,” giving developers a path from classical simulation to execution on real quantum hardware without committing to an exact release window. That caution stands out in an industry where premature launch dates routinely slip.
Inside Helios: The Quantum Computer Landing on OCI
Helios is Quantinuum’s flagship trapped-ion quantum processor, and the company markets it as the most accurate commercial quantum computer available today. Trapped-ion systems like Helios manipulate individual charged atoms with lasers to perform quantum operations, an approach that tends to produce higher gate fidelity than superconducting alternatives, at the cost of slower operation speeds. That tradeoff matters for enterprise buyers: fidelity determines whether a quantum result is trustworthy enough to act on, while speed determines whether it is practical to run production workloads.
Deploying Helios inside a U.S. Oracle AI data center, rather than requiring customers to reach out to a separate Quantinuum-hosted environment, is the structural part of this deal. It puts the quantum processor physically adjacent to the GPU clusters and storage that already run OCI’s AI workloads, cutting the data-movement friction that has made hybrid quantum-classical pipelines clumsy to build. Oracle frames this integration, not the qubit count or gate fidelity spec sheet, as the differentiator against Braket and Azure Quantum, both of which connect to external hardware partners over the public internet rather than co-locating processors inside the cloud’s own data centers.
Why Oracle Picked Quantinuum Over Rivals
Oracle didn’t need to build quantum hardware from scratch, and it didn’t need to sign every hardware vendor the way AWS did with Braket. Instead it picked one partner and went deep, mirroring the strategy Oracle has used elsewhere in its AI infrastructure buildout: fewer partners, tighter integration, faster time to a differentiated product. Quantinuum, for its part, gets distribution through a hyperscale cloud with an enterprise sales motion already selling database, ERP, and AI infrastructure contracts to the kind of large regulated customers (banks, pharmaceutical firms, logistics companies) most likely to fund early quantum experimentation.
Oracle CEO Safra Catz has been explicit that cloud growth, not just AI compute, is the company’s central bet for fiscal 2026. As she told Reuters during a quarterly results discussion, “We expect our total cloud growth rate — applications plus infrastructure — will increase from 24% in fiscal year 2025 to over 40% in fiscal year 2026.” A quantum computing exclusive gives Oracle a story that AWS, Microsoft, and Google can’t immediately match, at a moment when Catz has publicly committed to accelerating growth rather than just sustaining it.
The $143 Billion Quarter: Cloud Spending Hits an 8-Year High
Synergy Research Group’s Q2 2026 figures, cited by multiple outlets including The Register and CRN in early August, put worldwide enterprise spending on cloud infrastructure services at approximately $143 billion for the quarter, an increase of roughly 43% year over year. That is the fastest growth rate the market has posted in about eight years, driven largely by AI workload demand pushing customers to buy more compute, storage, and networking capacity than the underlying user growth alone would explain.
Market Share by Provider
The top three providers still control roughly two-thirds of the market, but the growth is broad enough that smaller providers are gaining ground too. Here is how Q2 2026 shook out, based on the Synergy-derived figures reported across multiple outlets:
| Provider | Q2 2026 Global Share | Notable Detail |
|---|---|---|
| Amazon Web Services | 28% | Still the single largest provider by spend |
| Microsoft Azure | 20% | Second-largest, driven by enterprise AI contracts |
| Google Cloud | 15% | Record-high share for Google, per Synergy |
| Top 3 combined | ~67% | AWS + Azure + Google Cloud |
| Oracle Cloud Infrastructure | Not disclosed publicly | Named among the fastest-growing tier-two providers, alongside CoreWeave, Crusoe, Nebius, and Nscale |
| Rest of market | ~33% | Split among Oracle, Alibaba, IBM Cloud, and dozens of regional and specialist providers |
Synergy’s full tier-two breakdown, including Oracle’s exact share, sits behind a paid research subscription, so the precise percentage isn’t public. What is public: analysts covering the Q2 numbers consistently place Oracle in the fastest-growing cohort of “tier-two” cloud providers, a group defined less by absolute market share and more by growth rate off a smaller base.
Where Oracle Fits Among Tier-Two Clouds
Oracle’s growth strategy over the past two years has leaned almost entirely on AI infrastructure contracts and specialized capacity, not general-purpose IaaS price competition against AWS or Azure. The Quantinuum partnership extends that pattern into quantum. Rather than trying to out-scale AWS on commodity compute, Oracle is stacking differentiated capabilities, AI superclusters, dedicated-cloud regions, and now exclusive quantum hardware access, to win specific enterprise workloads it couldn’t win on price or footprint alone.
Quantum Cloud Services Compared: AWS, Azure, Google, IBM, and Oracle
Oracle is not the first hyperscaler to offer quantum computing access. AWS launched Amazon Braket in 2020 as a managed marketplace for multiple quantum hardware vendors. Microsoft’s Azure Quantum followed a similar multi-vendor model. Google has taken a narrower, in-house approach built around its own superconducting processors and the open-source Cirq framework. IBM runs the largest fleet of cloud-accessible quantum computers of any single vendor. Oracle’s approach, one hardware partner deployed inside Oracle’s own data centers, is structurally different from all four.
| Cloud Provider | Quantum Service | Hardware Partners | Access Model |
|---|---|---|---|
| AWS | Amazon Braket | IonQ, Rigetti, IQM, QuEra, AQT | Per-task + per-shot pricing, plus hourly reservations |
| Microsoft Azure | Azure Quantum | IonQ (Forte 1, 36-qubit), Quantinuum H-series, historical Rigetti access | Per-job / per-shot pricing, device-specific |
| Google Cloud | Quantum AI / Cirq | In-house superconducting processors | Research collaboration access, no open pay-as-you-go tier |
| IBM Cloud | IBM Quantum | In-house superconducting processors (75 systems built, 13 cloud-accessible) | Pay-per-use plus IBM Quantum Network membership |
| Oracle Cloud Infrastructure | OCI quantum service (new, previewing soon) | Quantinuum (Helios) | Not yet disclosed; deployed inside Oracle’s own AI data center |
How Quantum Pricing Actually Works
AWS publishes the clearest public pricing of any hyperscaler quantum service, which is useful context for what Oracle will eventually have to disclose. Braket charges a flat per-task fee on top of a device-specific per-shot rate. IonQ’s Forte device runs $0.30 per task plus $0.08 per shot. QuEra’s Aquila neutral-atom device runs $0.30 plus $0.01 per shot. IQM’s Garnet processor runs $0.30 plus roughly $0.00145 per shot. Customers who need dedicated access can reserve hardware by the hour instead, at rates like $7,000 an hour for IonQ Forte or $4,000 an hour for IQM’s Emerald device.
A typical Braket job submission looks like this in the AWS SDK, which gives a sense of how a developer would eventually expect an OCI quantum job to be structured once Oracle publishes its own interface:
from braket.circuits import Circuit
from braket.aws import AwsDevice
device = AwsDevice("arn:aws:braket:::device/qpu/ionq/Forte")
circuit = Circuit().h(0).cnot(0, 1)
task = device.run(circuit, shots=1000)
result = task.result()
print(result.measurement_counts)Oracle hasn’t published an equivalent code sample or pricing sheet yet. Given the OCI quantum service is still in preview, expect Oracle to lean on its existing OCI SDK conventions rather than inventing a new developer surface from scratch.
What Oracle’s Executives and Partners Are Saying
Oracle’s own announcement kept the framing corporate and specific rather than promotional. “Quantinuum (NASDAQ: QNT), a leading quantum computing company, and Oracle today announced a multi-year strategic partnership to bring quantum computing to Oracle Cloud Infrastructure (OCI),” the company said in its official news release. Oracle Cloud’s own social account reinforced the workload framing directly: “We’re partnering with @QuantinuumQC to bring Helios to OCI and help customers explore hybrid quantum, AI, and HPC workloads,” the company posted the day of the announcement.
Quantinuum echoed the same commercial framing rather than a research-lab angle. “That’s why, today, Quantinuum and @Oracle have announced a multi-year strategic partnership to bring quantum computing to Oracle Cloud Infrastructure customers and accelerate the commercial adoption of hybrid quantum-AI computing,” the company wrote. Both companies are pitching this as a product launch aimed at paying enterprise customers, not a research pilot.
On the financial side, Catz’s Reuters comments about accelerating Oracle’s total cloud growth rate from 24% to over 40% year over year give the quantum deal a business rationale beyond novelty. A partnership like this costs Oracle relatively little to announce and gives sales teams a new pitch for large enterprise renewals, even before the service reaches general availability.
The Enterprise Readiness Gap: Ambition Outpaces Adoption
IBM’s Institute for Business Value published a Quantum Readiness Index in December 2025 that puts real numbers on the gap between corporate interest and corporate preparation. Fifty-nine percent of executives surveyed believe quantum-enabled AI will transform their industry by 2030, but only 27% expect their own organization to actually be using the technology by then. IBM’s researchers called that mismatch a strategic miscalculation, urging companies to start building quantum skills and governance now rather than waiting for the technology to mature on its own.
IBM CEO Arvind Krishna has made a similar point publicly, arguing that quantum’s practical payoff is still years out even as the hardware race accelerates. Speaking at IBM Think 2025, Krishna described a four-to-five year window before enterprises see real advantages from quantum systems, while noting IBM alone has already built 75 physical quantum computers and made 13 of them available on the cloud today. That combination, real hardware shipping now, meaningful returns still years away, is the backdrop every hyperscaler’s quantum pitch has to work against, including Oracle’s.
Historical Context: From Mainframes to Qubits in the Cloud
Cloud quantum access isn’t new. AWS Braket has offered managed access to multiple hardware vendors since 2020, and Azure Quantum followed a comparable multi-vendor model shortly after. What’s changed by 2026 is the scale of the classical infrastructure quantum services now sit next to. Oracle, AWS, and Microsoft have all spent the past two years building AI superclusters loaded with tens of thousands of GPUs. Bolting a quantum processor onto that same data center, instead of treating it as a standalone lab curiosity, reflects a bet that quantum’s near-term value comes from hybrid workflows: using classical GPUs for the bulk of a computation and routing only the specific subroutines that benefit from quantum mechanics to the QPU.
Microsoft has framed this shift in similar terms. CEO Satya Nadella told investors on a fiscal 2025 earnings call that quantum represents the next big accelerator for cloud computing, pointing to a deployment milestone with quantum computing company Atom Computing as evidence that Microsoft’s own hybrid strategy was moving from research toward production. Oracle’s Quantinuum deal follows the same logic roughly a year later, just with a different hardware partner and a tighter physical integration model.
Market Impact: What This Means for Enterprise IT Buyers
For most enterprise IT buyers, the immediate practical impact is small. Oracle hasn’t shipped pricing, hasn’t opened general availability, and hasn’t published which industries or workloads it expects to onboard first. What the deal does change is competitive pressure. Any large enterprise currently running a multi-cloud strategy across AWS, Azure, and Oracle now has a reason to ask Oracle’s sales team about quantum roadmap timing during contract renewals, even if actual usage is years away.
It also reframes how procurement teams should think about vendor lock-in. AWS and Azure’s multi-vendor hardware marketplaces let customers switch between IonQ, Rigetti, and other QPUs without leaving the cloud platform. Oracle’s single-partner model with Quantinuum offers less hardware flexibility, but potentially tighter integration and, if Oracle prices it competitively against Braket’s per-shot rates, a simpler bill. Buyers evaluating pilot quantum projects in 2026 and 2027 will need to weigh flexibility against integration depth, the same tradeoff that shows up in most single-vendor versus multi-vendor cloud decisions.
Competitive Response: How AWS, Microsoft, and Google Might Answer
AWS’s Braket marketplace model already covers Quantinuum’s ion-trap competitors, and it’s plausible Amazon could add Quantinuum hardware itself in the future given the marketplace’s open structure. Microsoft has already worked with Quantinuum’s H-series systems inside Azure Quantum, so an expanded partnership there wouldn’t require a new relationship, just deeper investment. Google remains the outlier: its quantum strategy stays research-focused and built around in-house hardware, without a Braket- or Azure Quantum-style open marketplace, which limits how quickly it can respond to a rival signing an exclusive hardware deal.
The more interesting competitive question is whether AWS or Microsoft follows Oracle’s co-location model, physically deploying a partner’s quantum processor inside their own AI data centers rather than routing traffic to an external facility. That would be a meaningfully larger infrastructure commitment than adding another vendor to an existing marketplace listing, and neither AWS nor Microsoft has announced plans to do so as of this writing.
Risks and Skepticism: Is This Hype or a Real Inflection Point
Skepticism about quantum cloud announcements is reasonable given how many “quantum advantage” claims from 2020 through 2024 didn’t hold up under scrutiny. Krishna’s own comments put a four-to-five year horizon on tangible enterprise benefit, and IBM’s readiness survey shows fewer than a third of executives expect to actually be running quantum workloads by 2030. Against that backdrop, Oracle’s announcement is a hardware access deal and an infrastructure commitment, not a claim of quantum advantage over classical computing for any specific business problem.
What makes this different from a pure marketing exercise is the physical infrastructure commitment: putting a live Helios system inside a production Oracle data center is a real capital and operational decision, not a slide in a keynote. Whether OCI customers actually adopt it at meaningful volume before 2028 is a separate question the preview period will start to answer.
Five Predictions for Cloud Quantum Through 2027
- Oracle will publish OCI quantum service pricing within two to three quarters of the August 2026 announcement, likely modeled on Braket’s per-task-plus-per-shot structure given how entrenched that pricing pattern already is across the industry.
- AWS or Microsoft will announce a comparable co-located quantum deployment (a QPU physically inside their own AI data center, not just marketplace access) within the next 12 to 18 months to blunt Oracle’s integration pitch.
- Early OCI quantum adopters will skew toward financial services and pharmaceutical firms already running Oracle database and ERP workloads, following the same customer base Oracle has used to sell its AI supercluster capacity.
- The gap IBM identified between executive interest (59%) and expected adoption (27%) will narrow only modestly by 2027, as most enterprises stay in pilot mode rather than production deployment.
- Synergy Research’s tier-two cloud growth numbers, the cohort that includes Oracle, CoreWeave, Crusoe, Nebius, and Nscale, will keep outpacing the top three hyperscalers’ growth rate through at least 2027, as AI and specialized-compute demand continues rewarding differentiated infrastructure bets over general-purpose scale.
Frequently Asked Questions
What did Oracle and Quantinuum actually announce?
A multi-year strategic partnership, announced August 11, 2026, to deploy Quantinuum’s Helios quantum computer inside a U.S.-based Oracle AI data center and offer access through a new OCI quantum service.
Can Oracle Cloud customers use quantum computing today?
Not yet. Oracle says the OCI quantum service will preview “in the coming months” but hasn’t published a firm launch date or pricing as of this writing.
How does Oracle’s quantum service compare to AWS Braket?
Braket has offered managed, multi-vendor quantum access (IonQ, Rigetti, IQM, QuEra, AQT) since 2020 with public per-task and per-shot pricing. Oracle’s model uses a single hardware partner, Quantinuum, physically deployed inside Oracle’s own data center, with pricing not yet disclosed.
What is Quantinuum’s Helios quantum computer?
Helios is Quantinuum’s flagship trapped-ion quantum processor, which the company markets as the most accurate commercial quantum computer currently available.
Why is global cloud spending growing so fast in 2026?
Synergy Research put Q2 2026 global cloud infrastructure spending at roughly $143 billion, up about 43% year over year and the fastest growth in about eight years, driven largely by AI workload demand across compute, storage, and networking.
Is quantum computing ready for enterprise production workloads?
Mostly not yet. IBM’s Quantum Readiness Index found 59% of executives expect quantum-enabled AI to transform their industry by 2030, but only 27% expect their own organization to be using it by then. IBM CEO Arvind Krishna has put the timeline for tangible enterprise advantage at four to five years out.
Which cloud provider has the most quantum hardware partners?
AWS Braket currently connects to the widest range of external hardware vendors, including IonQ, Rigetti, IQM, QuEra, and AQT. IBM runs the largest fleet of its own cloud-accessible quantum computers, with 13 systems currently available from a total of 75 built.
What does “hybrid quantum-AI computing” mean in this context?
It refers to workflows that combine classical GPU-based AI computing with quantum processing in the same pipeline, running the bulk of a workload on classical hardware and routing specific subroutines to a quantum processor, the model Oracle is pitching by integrating Helios with its existing OCI HPC and GPU infrastructure.
