Oracle to Deploy Quantinuum Quantum Computer in Its Data Center
▼ Summary
– Oracle and Quantinuum have partnered to install Quantinuum’s Helios quantum computer inside an Oracle Cloud Infrastructure data center, with Oracle selling access as a managed service; no financial terms, launch customer, or firm date were disclosed.
– Unlike typical brokerage models, Helios will be physically integrated with Oracle’s GPUs and HPC on the same network, enabling hybrid classical-quantum workloads under shared security controls.
– Helios, commercially launched in November 2025, is a trapped-ion machine with 98 physical qubits, 48 logical qubits, and 99.921% two-qubit gate fidelity, drawing about 60kW of power.
– Oracle enters the quantum cloud market late, after AWS, Microsoft, and IBM, but differentiates by offering deep integration of a single machine rather than a marketplace of remote quantum processors.
– Quantinuum’s 2025 Nasdaq IPO raised $1.68bn at a $14bn valuation on $30.9m revenue and a $192.6m net loss, with its roadmap targeting quantum advantage by 2029, while investors showed little reaction to the partnership.
Oracle and Quantinuum have entered a multi-year agreement that will see Quantinuum physically deploy its Helios quantum computer inside a U.S. Oracle Cloud Infrastructure (OCI) AI data center, with Oracle reselling access to the machine as a managed service. Financial terms were not disclosed, no anchor customer was named, and the only timeline given is a preview arriving “in the coming months.”
Most quantum cloud offerings today operate as a brokerage model. A user submits a job from a hyperscaler’s console, the task runs on a machine located at the quantum vendor’s own facility, and the results are returned. That distance matters. By placing Helios within an OCI data center, the quantum processor sits on the same network fabric as Oracle’s GPUs and high-performance computing clusters, governed by the same security and compliance controls. That proximity is what makes hybrid quantum-classical workloads viable, where a problem is shuttled between classical and quantum hardware multiple times during execution.
Helios itself is not new. Quantinuum launched the trapped-ion system commercially in November 2025, featuring 98 physical qubits, 48 demonstrated logical qubits, and an average two-qubit gate fidelity of 99.921%. The machine draws roughly 60kW of power, a figure that looks modest next to the tens of megawatts consumed by leading classical supercomputers, though that comparison flatters the system considerably given the narrow set of problems it can currently tackle.
Rajeeb Hazra, Quantinuum’s president and chief executive, framed the deployment as a chance to build something distinctive. “Deploying Helios inside OCI gives Quantinuum and Oracle an opportunity to create a unique deeply integrated environment for hybrid workloads, explore enterprise use cases with customers, and accelerate commercial adoption,” he said.
Oracle’s own language was more measured. Mahesh Thiagarajan, executive vice president of Oracle Cloud Infrastructure, said the goal is to give developers “a practical and secure way to explore how quantum computing could complement their existing AI and HPC workloads on Oracle Cloud Infrastructure while improving compute efficiency and energy use.” The named target applications follow the industry standard list: drug discovery, materials science, financial modeling, and large-scale optimization in logistics and energy.
Oracle is entering this arena late. AWS has offered Braket since August 2020. Microsoft’s Azure Quantum has been generally available since February 2022 and already carries Quantinuum’s H-series hardware. IBM has sold access to its own machines since 2016. Oracle’s pitch is not breadth but depth: one vendor, one machine, physically present, rather than a marketplace of remote quantum processors. Whether enterprises want that trade-off is precisely what the preview period is designed to test.
Notably absent from the announcement is Nvidia, despite Quantinuum’s existing NVQLink integration dating to last November and Oracle’s hybrid pitch relying heavily on GPU proximity.
For Quantinuum, this deal marks the second major validation in a year that has already transformed the company. The firm priced its Nasdaq IPO in June at $60 per share, selling 28 million Class A shares to raise $1.68 billion at a valuation near $14 billion, above its own upsized range and more than twenty times oversubscribed. That came on 2025 revenue of $30.9 million against a net loss of $192.6 million. The company trades under the ticker QNT, with a roadmap running to Sol in 2027 and Apollo in 2029, the machine it says will demonstrate genuine quantum advantage.
Europe has been pursuing a different model. Finland’s IQM has installed superconducting machines directly into research data centers rather than routing access through a hyperscaler, arguing that co-location with classical HPC is where useful work will first emerge. Oracle’s arrangement borrows that logic and applies it to a commercial cloud, which reads either as a sensible synthesis or a fast way to catch up, depending on your disposition.
The market reaction was muted. Oracle closed down 3.69% at $145.48 on Tuesday, while Quantinuum slipped 0.94% to $56.06 before recovering slightly in after-hours trading. A partnership with no disclosed terms, no named customer, and no launch date is difficult to price, which is roughly the position quantum computing has occupied for a decade.
What none of the parties has offered is a workload that demonstrably runs better this way today. The release leans on the phrase “explore enterprise use cases with customers,” the industry’s standard construction for work that has not yet produced results. Quantinuum’s own revenue of $30.9 million against a $192.6 million loss is a reasonable proxy for how much commercial quantum computing currently exists anywhere.
(Source: The Next Web)
