D-Wave

D-Wave (D-Wave Systems, headquartered in Burnaby, British Columbia, Canada) is a quantum-computing company and was the first to sell a commercial quantum computer, the D-Wave One, in 2011. Founded in 1999, its machines perform Quantum annealing rather than universal gate-model computation, so they target optimization problems and cannot run algorithms such as Shor's algorithm.

#Annealing, not a universal computer

A D-Wave processor is a quantum annealer built from superconducting flux qubits. It does not execute an arbitrary sequence of logic gates. Instead it encodes an optimization problem as an energy landscape, typically a quadratic unconstrained binary optimization (QUBO) or equivalent Ising model, and lets the hardware evolve toward a low-energy configuration that represents a good solution. This is a genuine and useful model of computation, but a restricted one. A gate-model Quantum computer running Shor's algorithm can factor integers and break RSA in principle; a quantum annealer cannot run Shor's algorithm at all. The two are different machines aimed at different problems, and conflating them is a common error.

#Hardware and topology

D-Wave has released a line of increasingly large processors. The D-Wave One (2011) used 128 qubits, followed by the 512-qubit D-Wave Two, the 1000-plus-qubit D-Wave 2X, and the 2000-qubit D-Wave 2000Q on the Chimera connectivity graph. The current generation, the Advantage system, uses more than 5000 qubits arranged in the Pegasus topology, giving each qubit up to 15 connections and allowing larger problems to be embedded directly. As of early 2026, D-Wave has also introduced Advantage2, which uses the denser Zephyr topology with 20-way qubit connectivity. Qubit count in an annealer is not directly comparable to gate-model qubit counts, because the qubits are specialized and never error-corrected.

#The quantum-speedup debate

D-Wave's early machines drew serious and sustained scientific debate on two questions: whether the device was doing anything genuinely quantum, and whether it solved problems faster than a good classical computer. Work led by Boixo and colleagues found behavior consistent with quantum annealing on the hardware (Nature Physics 2014). A separate study led by Ronnow and colleagues examined performance directly and reported no clear evidence of a quantum speedup over well-tuned classical solvers on the benchmarks tested (Science 2014). The consensus that emerged is that the presence of quantum effects and the existence of a practical speedup are separate claims, and that carefully optimized classical algorithms remained competitive on the problems studied.

More recently, D-Wave reported a beyond-classical result for simulating the dynamics of certain magnetic materials, arguing that no classical method could reproduce the computation in feasible time (King et al. 2024, published in Science in 2025). Several groups responded with classical tensor-network and other methods that they said narrowed or matched parts of the claimed advantage, so as of early 2026 the result is best described as a contested demonstration in quantum simulation rather than a settled speedup for a practical optimization task.

#Software and access

D-Wave provides the Ocean software development kit for building and submitting problems, and the Leap cloud service for running them on live annealers. Typical problem classes cast as QUBO or Ising models include scheduling, routing, and other combinatorial-optimization tasks, though whether an annealer beats a well-tuned classical solver on any specific instance is an empirical question rather than a guarantee. The company has also announced work on a gate-model quantum computer, which would be a distinct architecture from its annealing line. As of early 2026 the annealer remains its commercial product.

#Relation to cryptography

Because a quantum annealer cannot run Shor's algorithm, D-Wave's systems do not threaten public-key cryptography and are not relevant to Q-Day or to Bitcoin's quantum exposure. Claims that a D-Wave machine could break encryption confuse annealing with universal gate-model computing. The cryptographic threat model concerns fault-tolerant gate-model machines, a different technology from the annealers D-Wave sells.