BOULDER, CO — Quantinuum, the quantum computing firm established by a merger of Honeywell Quantum Solutions and Cambridge Quantum, announced the deployment of its third-generation quantum machine, Helios, significantly advancing computing power and precision through enhanced error correction. While the new system, based in Colorado, does not yet possess the capacity for full-scale commercial applications like financial modeling or materials discovery, its pioneering use of trapped barium ions as qubits positions the company at the forefront of scalable quantum architecture, setting it apart from competitors relying on superconducting circuits.
Helios represents a critical step on Quantinuum’s roadmap toward developing large-scale, fault-tolerant quantum systems. The machine operates using 98 barium qubits, a notable increase from the 56 ytterbium qubits utilized in its predecessor, H2. Barium ions offer improved controllability, making them easier to manipulate within the delicate quantum environment.
“Helios is an important proof point in our road map about how we’ll scale to larger physical systems,” stated Jennifer Strabley, vice president at Quantinuum, which remains majority-owned by Honeywell.
Precision and Scalability Drive Ion Trap Architecture
The core of Helios rests on a thumbnail-sized chip holding the barium ions, which are cooled to an extremely frigid 15 Kelvin (approximately -432 degrees Fahrenheit) inside a vacuum chamber featuring complex arrangements of lasers, mirrors, and optical fiber. Users can access the system remotely via cloud platforms.
Experts outside the company have lauded Helios for the inherent precision of its qubits. Rajibul Islam, a physicist at the University of Waterloo, remarked on the system’s initial low error rates, which minimize the hardware needed for error correction. Quantinuum reported a critical entanglement operation—the interaction between pairs of qubits—succeeded 99.921% of the time.
“To the best of my knowledge, no other platform is at this level,” Islam confirmed, highlighting the system’s unprecedented fidelity.
Beyond increasing the sheer number of qubits, a major technological breakthrough for Quantinuum is the ability to perform error correction “on the fly,” according to David Hayes, the company’s director of computational theory and design. This new capability employs Nvidia GPUs running in parallel to identify and mitigate errors in the stream of computation. Hayes noted that GPUs proved more effective for this real-time error management than the FPGAs (Field-Programmable Gate Arrays) commonly used in the industry.
Immediate Impact on Scientific Discovery
While full commercial profitability remains a distant goal for the quantum sector, Quantinuum is already leveraging Helios to conduct foundational scientific research. The company recently utilized the machine to simulate the complex behavior of electrons within a high-temperature superconductor, adding to prior work on H2, where it successfully simulated a magnet with complexity rivaling leading classical computing approaches. These simulations offer physicists new insights into phenomena like magnetism and superconductivity.
The company is scaling its infrastructure, planning to install a new Helios system in Minnesota. Looking ahead, Quantinuum outlined an aggressive timeline for future generations:
- Sol (2027): Anticipated fourth-generation system with 192 qubits.
- Apollo (2029): Expected fifth-generation architecture designed to incorporate thousands of qubits and achieve full fault tolerance.
The successful introduction of Helios reinforces the potential of ion-trap quantum computing as a viable, scalable alternative to superconducting architectures, accelerating the industry’s progress toward delivering practical, commercially useful quantum power.
