As the year winds down, quantum computing often sees a wave of announcements as companies strive to meet end-of-year commitments. This year, however, includes a significant development: the debut of dual-rail qubit technology, introduced by Quantum Circuits. This new approach aims to make error detection and correction more efficient by addressing the most common quantum computing error—photon loss. Although tech giant Amazon has explored dual-rail qubits, Quantum Circuits has become the first to make them publicly accessible via a cloud service.
Dual-rail qubits represent a new variation on transmon hardware, widely used by companies like Google and IBM. The system employs two linked superconducting setups that allow microwave photons to move between them. This enables the storage of quantum information in the probabilistic location of a photon. While the setup requires twice the hardware per qubit compared to traditional designs, it significantly simplifies the detection of photon loss errors, which constitute the majority of errors in quantum systems.
Photon loss, accounting for over 90% of quantum computing errors, is straightforward to detect with dual-rail qubits. The system allows for non-disruptive checks of photon integrity, signaling an error if the expected photon count is incorrect. While this is invaluable for simpler computations, it does not eliminate the need for broader error-correction techniques to handle rarer errors like phase flips. Additionally, even when photon loss is detected, additional measurements are required to determine how to correct the error effectively.
Quantum Circuits’ initial hardware is intentionally limited to eight dual-rail qubits, focusing on error correction rather than large-scale computation. This design is intended to help users experiment with and learn the unique capabilities of dual-rail qubits, especially in error detection. To support this, Quantum Circuits developed its own software stack tailored to the hardware’s features and chose to host its system on a proprietary cloud platform to directly engage with early adopters and refine the technology based on user feedback.
Despite entering a competitive field where companies like IBM and Google have achieved systems with hundreds of qubits, Quantum Circuits and similar startups are focusing on hardware with fewer qubits. This may seem counterintuitive, but it reflects a strategic focus on error correction, widely regarded as the cornerstone of future quantum computing. Rather than emphasizing qubit count, these companies aim to address error rates to build scalable and reliable quantum systems.
The dual-rail qubits illustrate how tackling error rates is more complex than scaling qubit numbers. Scaling primarily involves replicating proven designs, while reducing error rates requires solving deep physical challenges, such as improving the materials used for superconducting circuits or refining control systems to minimize fluctuations. A lower error rate reduces the number of hardware qubits needed to create logical qubits, making large-scale quantum computing more achievable.
Established companies have similarly prioritized error correction. Google has iteratively improved its chip designs since its 2019 quantum supremacy milestone, focusing on understanding and reducing error sources. IBM has shifted from scaling to refining smaller processors with lower error rates. This shared focus highlights a common realization: scaling hardware without addressing errors offers limited benefits, as high error rates undermine computational reliability.
New entrants like Quantum Circuits are banking on novel approaches to error correction to distinguish themselves. Technologies like dual-rail qubits simplify error detection, while companies like Oxford Ionics and Alice & Bob introduce methods to minimize or manage errors differently. These startups aim to establish their technology as a foundation for scalable systems, competing in a market still full of opportunities despite the dominance of established players with greater resources and experience.
