Quantinuum and Oracle have announced a multi-year strategic partnership that will place Quantinuum's third-generation quantum computer, Helios, inside a US-based Oracle Cloud Infrastructure (OCI) AI data center. The move is designed to give OCI customers managed access to quantum computing alongside Oracle's existing GPU and high-performance computing resources, with the goal of supporting hybrid quantum-AI workloads.

Under the agreement announced August 11, 2026, OCI customers will be able to access Helios through a planned OCI quantum service, rather than procuring or operating dedicated quantum hardware. The companies said the integration is intended to connect the quantum system with OCI's compute, networking, storage, identity, and data services under the same governance and access controls customers already use.

What's New

Helios is Quantinuum's 98-physical-qubit trapped-ion system, launched commercially in November 2025. It has been used in demonstrations involving 48 logical qubits and achieves an average two-qubit gate fidelity of 99.921%, exceeding the widely cited "three 9s" threshold. The system uses barium-137 ions as physical qubits, with ions moved between interaction zones where lasers perform quantum operations and measurements.

Oracle plans to preview its OCI quantum service in the coming months. The service is expected to combine Quantinuum's development stack with support for open-source hybrid-programming frameworks, giving developers a path from simulating quantum circuits on classical hardware to executing them on the real machine in the same environment.

Quantinuum has also integrated Helios with GPU-based classical computing for real-time quantum error correction. Measurement data from the quantum processor is decoded by classical hardware before correction information is returned during the computation. The Helios product page lists single-qubit gate fidelity at 99.9975% and notes that NVIDIA Grace Hopper GPUs sit inside the control system itself.

Why It Matters

The partnership reflects a shared vision that the future of enterprise computing will be built on the convergence of AI, classical supercomputing, and quantum computing. Many complex problems across materials discovery, drug development, logistics, energy, and financial modeling already push the limits of today's computing architectures.

The energy comparison is a notable part of the pitch. Citing a 2026 academic survey of energy-aware computing, the companies put leading supercomputers at 16 to 39 megawatts of draw against roughly 60 kilowatts for a single Helios unit without its HVAC system, less than one percent of a top supercomputer's consumption. Quantinuum's own product materials rate the base Helios unit at under 40 kilowatts.

That comparison comes with an obvious caveat: Helios is a complementary resource for suitable hybrid workloads, not a replacement for the GPU clusters OCI is building at gigawatt scale. The release itself describes quantum computing as addressing problems "impractical for traditional systems alone," which is the accurate frame.

For hybrid algorithms that hand specific subroutines to a quantum processor, offloading those steps to a machine drawing kilowatts rather than running them classically at megawatt scale is the efficiency claim. It is why Oracle is presenting the deployment as part of its infrastructure story rather than a research sideshow.

Our Take

The Oracle deal lands on a busy day for Quantinuum, which also reported its second-quarter 2026 results on August 11, its first formal guidance as a public company. Revenue was $8 million for the quarter ended June 30, up 279% year over year, with full-year 2026 revenue expected between $28 million and $32 million. The company completed what it describes as the quantum industry's first traditional IPO earlier in 2026, raising $1.7 billion in gross proceeds, and held $2.1 billion in cash and short-term investments as of June 30.

The deployment model follows a pattern Quantinuum has been building: enterprise and national-infrastructure installations over pure cloud brokering. Its roadmap calls for the Sol system in 2027, with the trap chip back from fabrication and moving through product validation, and the fault-tolerant Apollo system in 2029. The Oracle arrangement extends the same strategy of putting hardware where enterprise workloads already live.

What ships next is the OCI quantum service preview, which is when developers get the first concrete look at how simulation-to-execution workflow, pricing, and access tiers will work. The release carries the customary Oracle disclaimer that development, release, timing, and pricing of described features remain at Oracle's sole discretion, so the preview is the event that turns this announcement into a product.

Sources