Oracle And Quantinuum Show Where Enterprise Quantum Is Heading

Putting a quantum computer inside a cloud data centre sounds like the interesting part of Oracle and Quantinuum's new partnership. It isn't. On August 11, the companies announced that Quantinuum's Helios quantum computer will be deployed inside a US-based Oracle Cloud Infrastructure (OCI) AI data centre. 

Customers will eventually be able to access Helios through Oracle's planned quantum service and combine it with OCI's existing GPU and high-performance computing (HPC) infrastructure. Oracle says the service will be previewed in the coming months, although there's no confirmed deployment date yet.

Cloud access to quantum computers has been around for years. What's more interesting is the infrastructure Oracle and Quantinuum are building around Helios. Because enterprise quantum computing may not arrive as a completely separate technology stack after all.

Quantum Computing In The Cloud Isn't The New Part

Amazon Braket already gives developers access to quantum processing units (QPUs) from several hardware providers, including QuEra, IonQ, IQM, Rigetti and AQT. AWS also provides simulators and tools for developing quantum applications without organisations having to own or operate the underlying machines themselves.

So Oracle isn't breaking new ground simply by making a quantum computer available remotely. The difference is how Helios is expected to fit into OCI. Quantinuum says the system will operate inside Oracle's infrastructure and is expected to integrate with existing OCI compute, networking, storage, identity and data services. 

Customers should also be able to manage it through the same governance and access controls they already use across OCI. That shifts the conversation from access to integration.

Instead of sending an experimental workload off to a separate quantum environment, an organisation could eventually have a QPU sitting alongside the classical computing resources already supporting the application. And once you look at quantum computing that way, the architecture starts to feel surprisingly familiar.

Quantum Is Starting To Look More Like An Accelerator

Modern computing environments already use different processors for different jobs. A central processing unit (CPU) remains the general-purpose workhorse. A graphics processing unit (GPU) is much better suited to running large numbers of calculations in parallel, which is one reason GPUs have become so important to AI. 

Other specialised processors are added when particular workloads need them. A quantum processing unit could eventually fit into the same basic model. That doesn't mean replacing CPUs, GPUs or conventional supercomputers with quantum machines. 

Quantum computers work differently and are only expected to offer an advantage for particular types of problems. Much of an application would still run on classical infrastructure even if one part of its workload eventually benefited from quantum computing. 

So the more useful distinction isn't classical computing versus quantum computing. It's classical computing plus quantum computing. Oracle and Quantinuum are explicitly building towards that idea. 

Their stated vision brings quantum computing together with AI and HPC rather than treating the technologies as competing alternatives. Helios itself was designed for hybrid integration with classical infrastructure. The QPU becomes another specialist the application can call when it has the right job for it. And Oracle isn't the only company heading in that direction.

Oracle Isn't The Only Company Building Towards This Model

Quantinuum has already been working with NVIDIA on integrating quantum and accelerated computing. Helios uses an NVIDIA Grace Hopper system as a real-time classical host, while NVIDIA's NVQLink architecture is designed to connect quantum processors with GPU-based computing infrastructure.

AWS and QuEra are approaching the problem differently, but there's a similar architectural idea underneath it. In June 2026, they announced that QuEra's planned Libra fault-tolerant quantum computer will become available through Amazon Braket in 2028. AWS describes its wider direction as bringing quantum computing into an accelerated computing portfolio where quantum, HPC and AI resources can support hybrid workflows.

These systems aren't interchangeable. Their hardware, maturity, technical approaches and timelines differ considerably. But the direction is becoming easier to see. Hybrid quantum-classical computing increasingly assumes that a QPU will exist inside a larger computing environment rather than operating as an island outside it.

That could change what enterprise quantum adoption eventually looks like.

Enterprise Quantum Adoption May Look More Familiar Than Expected

Quantum adoption is often described as though organisations will eventually reach a point where they have to decide whether they're "going quantum". The integrated cloud model suggests something less dramatic. Imagine an organisation already running applications, data and AI workloads in a cloud environment. 

Its developers don't build and operate the GPUs underneath every AI service they use. Infrastructure teams don't need to understand the physical design of every processor either. They consume those resources through services, APIs and platforms, while the cloud provider deals with much of the underlying hardware.

Quantum could eventually follow a similar path. A developer might build a hybrid quantum application that sends one part of a calculation to a QPU while the rest remains on conventional infrastructure. Identity, networking, storage and access could still be governed through the organisation's existing cloud environment. 

Specialist quantum expertise would remain important for deciding whether a problem actually suits quantum computing, but enterprises wouldn't necessarily need an entirely separate operating model simply to use it. That reframes one of the biggest questions around enterprise quantum computing.

Instead of asking, When should we adopt quantum computing?, technology leaders may eventually be asking: When does this particular workload justify using the quantum capability available inside our existing computing environment? We're nowhere near the point where most organisations need an answer to that question. 

But cloud providers are beginning to build the infrastructure that would make asking it practical.

The Infrastructure May Arrive Before The Business Case

This is where a little restraint becomes useful. Oracle hasn't announced when Helios will actually be deployed in OCI. Its quantum service is only expected to enter preview in the coming months. 

And while Helios is already a commercial system, putting a QPU inside enterprise cloud infrastructure doesn't prove that ordinary enterprise workloads will suddenly perform better on it. 

Quantum computing still has significant technical hurdles to clear, including error correction, fault tolerance and demonstrating useful advantages over increasingly powerful classical systems. 

Our previous analysis of enterprise quantum readiness has made the same distinction: organisations can prepare for quantum without pretending broad production adoption has already arrived. The phrase quantum AI needs similar care. Oracle and Quantinuum are talking about hybrid quantum-AI infrastructure. 

That doesn't mean enterprise AI workloads are about to move wholesale onto quantum processors. It means they're building an environment where quantum, AI and HPC resources can potentially be combined when a useful workload emerges. Infrastructure can mature before the business case does. In fact, it probably has to.

Developers need programming frameworks. Infrastructure teams need orchestration and governance. Hardware needs reliable integration with classical systems. Organisations need to understand which workloads are even worth testing. The important distinction for enterprise leaders is therefore between infrastructure readiness and workload readiness.

One can move considerably faster than the other.

Final Thoughts: Quantum May Arrive As Part Of The Stack

Oracle putting Helios inside OCI doesn't mean mainstream enterprise quantum computing has arrived. It does give us a more useful picture of how it might eventually get there. The interesting shift isn't putting quantum computing "in the cloud". 

It's designing quantum capability into the same computing environments already running enterprise applications, AI and HPC workloads. If that model continues to develop, organisations may not experience quantum adoption as one enormous technology change. 

A QPU could simply become another specialised computing resource that's used when a workload finally gives it something worth doing. That means qubit records and hardware announcements only tell us part of the story. 

Workload integration, orchestration, developer tooling, fault tolerance and demonstrated quantum advantage will be better signs that enterprise quantum is becoming genuinely useful. For now, the infrastructure appears to be arriving ahead of many of the workloads.

EM360Tech will keep following that gap as cloud providers, quantum companies and enterprise developers begin working out where quantum computing actually belongs in the technology stack.