D-Wave is changing its approach to quantum computing. While the company built its reputation on quantum annealers, it is now advancing into gate-based hardware using dual-rail qubits. This technology represents a significant shift in how the company manages error correction within complex quantum systems.
The core of this development involves a two-qubit gate design that the company recently validated in the journal Nature. Dual-rail qubits work by using two linked resonators to hold a single photon. This structure makes photon loss the most common type of error. Because loss is easy to detect, researchers can avoid the heavy overhead of traditional error-correction methods that require extra qubits for every calculation.
During testing, D-Wave successfully entangled two dual-rail qubits in 500 nanoseconds. This speed is critical for practical applications. The team maintained an error hierarchy where photon loss remained the dominant issue, confirming that the entanglement process does not introduce chaotic new error types. By keeping bit-flip errors at the 10-6 level, the system preserves the benefits of the dual-rail design.
There are still hurdles ahead for the team. The experimental devices showed an unexpected decline in performance as the number of operations increased, likely due to calibration drift. Additionally, the company needs to perfect mid-circuit erasure detection to fully realize the potential of this architecture. D-Wave intends to scale this hardware to 181 dual-rail qubits by 2028.
This transition marks a new phase for the organization as it aims to host more than a hundred logical, error-corrected qubits in future iterations. By focusing on simplified error codes, the company believes it can deliver high-performance computation with less hardware than competing gate-based designs. The success of this dual-rail roadmap will depend on how effectively the team manages the scaling of these gates.