D-Wave Quantum said it is targeting a 100-logical-qubit gate-model system by 2032 that it says will be capable of more than 1 million successful operations, and laid out a step-by-step roadmap extending from 2026 through 2032.
The company’s plan starts with a 17-physical-qubit system in 2026 that it says will cut logical error rates to half the physical error rate, then a 49-physical-qubit system in 2027 with an expected 20-fold error reduction, and a 181-physical-qubit system in 2028 with an expected 2,000-fold error reduction. By 2030, D-Wave is aiming for a 10-logical-qubit system that can support fault-tolerant algorithms, before reaching the 100-logical-qubit milestone in 2032.
D-Wave said its dual-rail architecture is designed to detect about 90% of errors as they occur, and said it has already demonstrated 99.9% two-qubit fidelities with error detection, meaning about one physical error in every 1,000 operations.
The company said its roadmap is built around a “lambda” metric, with the broader industry at around 2 and D-Wave targeting 10. On that basis, D-Wave said each added layer of error correction would reduce errors by a factor of 10, rather than roughly half.
D-Wave also said its superconducting technology can run quantum error correction cycles 100 to 1,000 times faster than neutral atom or trapped ion systems.
The roadmap was presented alongside D-Wave’s first investor day, and the company said it has more than 15 years of experience building superconducting quantum systems and has delivered six generations of annealing quantum computers, including its Advantage2 system. The market has reacted to these announcements by moving the company's shares -0.17% to a price of $30.09. For the full picture, make sure to review D-Wave Quantum's 8-K report.
