Quantinuum advances hybrid quantum computing with Helios launch and global partnerships

Quantinuum, the quantum computing company formed from the merger of Honeywell Quantum Solutions and Cambridge Quantum, is pushing deeper into the enterprise market with its Helios quantum computing platform.

Helios, which Quantinuum launched on November 5, 2025, packs 98 qubits using a trapped-ion architecture. Its two-qubit gate fidelity sits at 99.921%, a number that matters enormously in a field where tiny errors compound fast enough to make calculations useless.

What Helios actually does

To understand why gate fidelity matters, think of it like a game of telephone. Each quantum operation is a whisper passed between players. At 99% fidelity, you lose 1% of accuracy per step. String enough steps together and your answer is garbage. At 99.921%, those errors shrink dramatically, meaning Helios can run longer, more complex calculations before noise overwhelms the signal.

The system is designed specifically to support generative quantum AI applications. It integrates with NVIDIA GB200 processors to enable hybrid quantum-classical workflows, essentially letting quantum and traditional computers team up on problems that neither could solve alone efficiently.

The partnership landscape

Quantinuum has lined up an impressive roster of early collaborators for Helios. Amgen, BMW Group, JPMorgan Chase, and SoftBank were all named as part of the launch. Each represents a different slice of the enterprise market: pharmaceuticals, automotive manufacturing, financial services, and telecom/venture capital.

Perhaps the most strategically significant partnership came on the same day as the Helios launch. Quantinuum inked a deal with Singapore’s National Quantum Office that includes plans to install a Helios system in the country by 2026 and establish local R&D operations.

The cloud quantum race

Oracle Cloud has been developing quantum-related capabilities including quantum simulation tools and post-quantum cryptography solutions. Oracle has worked with companies like Arqit and American Binary on post-quantum security, and has integrated NVIDIA cuQuantum appliances for quantum AI workloads. These efforts are independent of Quantinuum and the Helios platform.

The cloud delivery model is critical because quantum computers are extraordinarily finicky machines. Trapped-ion systems like Helios require precise laser control and ultra-high vacuum environments. Superconducting systems from competitors need temperatures colder than outer space.

What differentiates Quantinuum is the combination of high fidelity and the trapped-ion approach. The 99.921% two-qubit gate fidelity Helios achieves is among the best publicly reported figures in the industry, and fidelity improvements have an outsized impact on the practical usefulness of quantum machines.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.