Quantinuum

Quantinuum is a quantum-computing company formed in 2021 by combining Honeywell Quantum Solutions with Cambridge Quantum. It builds gate-model quantum computers from trapped ions in its H-series and is known for very high gate fidelities and for repeatedly setting Quantum Volume records.

#Origins

The company was created when Honeywell's trapped-ion hardware group merged with Cambridge Quantum, a quantum-software firm. This gave Quantinuum both a hardware line and a software stack under one organization. Cambridge Quantum, based in the United Kingdom, brought software spanning quantum chemistry, machine learning, and cybersecurity, including a hardware-based quantum-random-number product. Honeywell remained the majority owner following the merger. The combination is sometimes described as a full-stack company, spanning ion-trap hardware, error correction, and application software.

#H-series hardware

Quantinuum's processors are the H-series, including the H1 and H2 systems. They use a quantum charge-coupled device (QCCD) architecture, in which ions are physically shuttled between zones of a segmented trap to bring arbitrary pairs together for gates, providing effective all-to-all connectivity. The H2 processor uses a race-track shaped trap and, as reported by the company and in the literature, achieved high two-qubit gate fidelities and record Quantum Volume figures (Moses et al. 2023). Quantum Volume is a single-number benchmark, and Quantinuum's repeated records on it reflect the modality's combination of high fidelity and flexible connectivity rather than a large qubit count. The QCCD design also supports mid-circuit measurement and qubit reuse, in which a qubit is measured partway through a circuit and then reset and used again, features that matter for error correction. As of early 2026 the company has also described a newer, larger-scale system in its roadmap.

#Error correction and logical qubits

Quantinuum's high physical fidelities have made its hardware a frequent testbed for Quantum error correction. In 2024, work carried out with Microsoft on H2 reported the creation of several logical qubits with error rates lower than the underlying physical qubits, along with repeated rounds of error correction (Microsoft and Quantinuum 2024). These were small-scale demonstrations of encoded qubits and logical operations, not a fault-tolerant computer, but they are among the stronger experimental results on encoded qubits reported to date. Quantinuum also develops the TKET compiler, which optimizes and translates circuits for different hardware backends, along with other software used across its stack. Because the same processors serve both the hardware and the error-correction research, results reported on H-series machines are often used as reference points for logical-qubit performance across the field.

#Relation to cryptography

Quantinuum builds universal gate-model machines, so a sufficiently large and error-corrected trapped-ion system could in principle run Shor's algorithm. Current H-series devices are far too small for that, with tens of physical qubits and only a handful of logical qubits demonstrated. The company's strong error-correction results are relevant to the long-run path toward a cryptographically relevant quantum computer, but that path remains a multi-year scaling problem shared by every hardware platform.