H-Series
When developing its H-Series quantum computers, Powered by Honeywell, Quantinuum chose a quantum charge-coupled device (QCCD) architecture as its path to scalable universal quantum computing because it allows full connectivity between identical high-fidelity qubits (atomic ions).[3]
Quantinuum launched its first generation of quantum computers with the System Model H1-1, a trapped-ion computer running on 12 qubits, in 2020.[10]
In May 2023, Quantinuum launched the System Model H2, with a quantum volume of 65,536 (216), the largest on record at that time. The H2 achieved the largest GHZ state on record, the first demonstration of magic state distillation, and the first demonstration of the creation and control of topological qubits whose linking properties can help make quantum computing fault-tolerant.[11][12] Braiding quasiparticles called non-Abelian anyons creates a historical record of the event, and the paths they trace are more robust to errors, which could eventually lead to the development of a topological quantum computer.[11]
The H-Series systems have consistently broken records for quantum volume, recently reaching a quantum volume of 33,554,432 (225) in September 2025.[2] Quantum volume is one of 15 performance benchmarks that Quantinuum scientists measured on the latest generation of its trapped-ion quantum computer, System Model H2.[13]
The company also holds the record for two-qubit gate fidelity, becoming the first to reach 99.9%. Microsoft and Quantinuum created four logical qubits on the H2 quantum computer, running 14,000 experiments without a single error.[14][15]
Demonstrating the ability to scale the H-Series, the company solved the qubit “wiring problem,” using a new chip arranged in a 2D grid to efficiently sort qubits and minimize the number of signals required for qubit control. The work reduces the ratio from as many as 20 analog wires per qubit to 1 digital wire per qubit with a fixed number of analog lines.[16][17]
Quantinuum also offers an H-Series Emulator, which allows researchers to compare data in quantum hardware experiments and approximate noise, accelerating simulation workflows.[12]
Quantum cybersecurity – Quantum Origin
Quantum Origin uses quantum computing to strengthen the cryptographic keys that protect online transactions and identification processes. The software produces provably unpredictable cryptographic keys to support traditional algorithms, such as RSA and AES, as well as post-quantum cryptography algorithms.[4][18][19][20][21][22]
Quantum Origin is said to be the first commercial application of a quantum computer offering a solution that classical computers cannot achieve.[23]
InQuanto is a quantum computational chemistry software platform. InQuanto uses Quantinuum's open-source Python toolkit, TKET, to improve the performance of quantum devices with electronic structure simulations. The stand-alone platform is designed to help computational chemists experiment with quantum algorithms and eventually create prototypes of real-life problems using quantum computers.[31]
TKET is a platform-agnostic compiler for optimizing quantum algorithms as well as a software development kit for building and running programs for gate-based quantum computers. It is platform-inclusive and open source. The quantum programming environment is accessible through the PyTKET Python package, with extension modules that work with quantum computers, classical simulators, and quantum software libraries.[32][33][34][35][36][37]
Quantum NLP/Compositional Intelligence
Quantinuum's Quantum Natural Language Processing team is developing reasoning-based quantum artificial intelligence that works with modern machine learning-based techniques to produce AI systems that are more interpretable, transparent, and cost effective, requiring less data.[38][39] This quantum compositional intelligence is based on categorical quantum mechanics, which studies quantum processes and how they are composed.
Quantum Monte Carlo Integration
Quantinuum's full Quantum Monte Carlo Integration engine is designed to use quantum algorithms to perform estimations more efficiently and accurately than equivalent classical tools, inferring an early-stage quantum advantage in areas such as derivative pricing, portfolio risk calculations and regulatory reporting.[40][41]
Lambeq
Lambeq is an open-source software library for the design and implementation of quantum natural language processing applications.[42] To build a quantum natural language processing model, Lambeq parses the grammatical structure of an input sentence into a task-specific output. This is encoded into an abstract representation called a string diagram, which reflects the relationships between the words in the original sentence.[42]
Quantum machine learning
Quantinuum has efforts in QML with a focus on quantum circuit learning on near-term noisy intermediate-scale quantum (NISQ) computers. The company has commercial work in deploying deep learning for time-series modeling and decision-making and specializes in quantum enhanced solutions for machine learning and optimization problems.[43][44]
At the intersection of classical machine learning and quantum computing, Quantinuum collaborated with Google DeepMind to use AI (Alpha-Tensor) to optimize the T-gate count. This research serves to minimize the compute costs resulting from one of the most expensive quantum logic gates in terms of time and resources required.[45]
Optimization
Among the primary uses for quantum computing is combinatorial optimization, as its applications extend to logistics, supply chain optimization, and route planning.
In 2023, Quantinuum created an improved variational quantum algorithm for solving combinatorial optimization problems that uses minimal quantum resources and takes advantage of the H-Series’ all-to-all connectivity and native parameterized two-qubit gates.[46] In 2021, Deutsche Bahn partnered with Quantinuum to explore how quantum computers can improve the rescheduling of rail traffic.[47]
Simulation
In 2021, Nippon Steel Corporation used Quantinuum's algorithms to simulate the behavior of iron crystals in different configurations. The chemical simulation is so complex at scale that it cannot be accurately simulated on classical computers.[48]
