In the wake of a planned merger announcement for Honeywell Quantum Solutions (HQS) and Cambridge Quantum (CQ), the pair jointly made three significant quantum announcements this week. The announcements covered a lot of ground – quantum error correction, a new quantum volume record, and speedup of important quantum optimization software. HQS and CQ have a long history of working together on quantum projects that involve optimization, scheduling, and other enterprise-level challenges.

Honeywell Quantum Solutions began in 2018 as a business unit of Honeywell International. Its quantum hardware technology of choice is trapped ytterbium ions. Cambridge Quantum (CQ) was founded in 2014 and develops quantum software for quantum chemistry, quantum machine learning, and quantum augmented cybersecurity. The merger should close in Q3 2021

**A big step in error correction**

Developing scalable error correction is key to the future of quantum computing. Unlike classical computers, which rarely make errors, quantum computers are error-prone and susceptible to errors caused by environmental factors such as background radiation and noise generated by cabling and electrical components. Qubits, the fundamental computational units for quantum computers, also contribute to errors in other qubits.

The inability to correct errors at scale is one of the main reasons we cannot build large, fault-tolerant quantum computers today. There has been a great deal of research in error correction. These schemes all require the use of some number of physical qubits to produce one good computational qubit. Estimates for the number of physical qubits needed to create one logical qubit vary from five to a thousand or more. For example, 100 logical qubits would require 100,000 physical qubits for error correction at the high end. For really useful work, like large molecule simulation, it would need millions of physical qubits.

Honeywell’s Model H1 quantum computer pioneered the use of QCCD, an advanced trapped-ion architecture that allows for arbitrary movement of ions and parallel gate operations across multiple zones.

Taking advantage of its QCCD architecture, for a first in error correction, Honeywell researchers demonstrated * repeated* rounds of real-time quantum error correction. This research represents a significant step toward the realization of large-scale quantum computing.

**System Model H1 doubles its Quantum Volume (again)**

This week HQS announced it had achieved a *measured* quantum volume of 1,024, the world’s highest measured quantum volume. quantum volume measures a quantum system’s overall performance. Adding qubits alone cannot increase a quantum computer’s power. Its performance is affected by its architecture and the interaction of the number of qubits, connectivity of qubits, gate fidelity, cross talk, and circuit compiler efficiency. Quantum volume is a good measurement because it considers all those factors.

Honeywell has met its forecasted annual doubling of quantum volume (QV) objectives by making continual improvements in the architecture and hardware of its Model H1 quantum computer. The previous measured QV records also achieved by Honeywell were:

- March 2021 Model H1 achieves QV of 512
- September 2020 Model H1 achieves QV of 128
- June 2020 Model H0 achieves QV of 64
- March 2020 first Honeywell Model H0 quantum computer announced

**A new VQE-type quantum algorithm**

VQE algorithms estimate the lowest energy state or a system’s minima. VQE is essentially a hybrid machine learning algorithm that splits the work between classical computers and quantum computers. It minimizes cost functions to determine the best way to load several vehicles, the most efficient vehicle routing, or supply chain costs. It is a complex process that requires many iterations.

Cambridge Quantum announced it has created a proprietary quantum algorithm that needs far fewer qubits to solve optimization problems. The algorithm also uses new methods to accelerate convergence up to 100 times faster, improve the solution quality, and reduce hardware resource requirements. This algorithm demonstrated it performed better than the original VQE algorithm and the Quantum Approximate Optimization Algorithm (QAOA). These new methods were developed and implemented on Honeywell’s System Model H1 quantum processor.

According to the joint press release by HQS and CQC, Ilyas Khan, CEO, and founder of Cambridge Quantum, provided context to the announcements. “Faster quantum algorithms can have a profound impact on a variety of industries that face complicated optimization problems. An excellent example is steel manufacturing, where global steel companies typically produce a variety of products. Manufacturing at this scale and complexity on time at minimal cost requires complicated scheduling of several production processes that are challenging for even the largest classical computing systems currently available. Logistics companies, airlines, and in a slightly different context, diversified financial services companies and banks need the same type of solution. By optimizing these processes, companies and, ultimately, its customers and consumers, in general, can see the positive effects. Honeywell and Cambridge Quantum are making it easier for businesses to do its jobs well and effectively.”

**Analyst notes:**

- Honeywell’s error correction technique is significant because it demonstrates the ability to repeat error correction cycles and determine needed corrections during computation. The logical qubit can be repeatedly error corrected. This is a crucial feature needed for future quantum computer scaling. Error correction protocols prior to this method were not repeatable.
- Honeywell’s repeatable error correction requires a quantum processor to have mid-circuit measurement and reuse. At present, only IBM and Honeywell have that capability.
- Choosing a benchmarking protocol depends on if you are interested in benchmarking all or part of a quantum computer. There are protocols for testing low-level quantum devices for fabrication or testing subsystems like randomized benchmarking. Quantum volume measures the entire system.
- Cambridge Quantum has made combinatorial optimization more efficient using its new Filtering Variational Quantum Eigensolver (F-VQE), which utilizes filtering operators to achieve faster and more reliable convergence to the optimal solution. You can read its research paper on F-VQE here.
- I am still hoping to hear more announcements from Honeywell. It has the technical capability to add additional ions to its QCCD trap. IonQ already has 30 ions in a linear trap with the ability to add more. However, it makes sense for Honeywell to hold that announcement either for post-merger or, if it goes public, just after that announcement.

*Note: Moor Insights & Strategy writers and editors may have contributed to this article.*