D-Wave Quantum

Overview

D-Wave takes a different approach from companies such as IBM, Google, and Quantinuum. Its machines are quantum annealers: special-purpose devices that look for the lowest-energy solution to an optimization problem by letting a network of superconducting qubits settle into a low-energy state. They cannot run general-purpose algorithms such as Shor's, but they can address scheduling, routing, and sampling problems that map naturally onto an energy landscape.

For years that focus made D-Wave unusual and sometimes controversial. Today the company is trying to broaden its position. It continues to sell and cloud-host annealing systems, and since the acquisition of Quantum Circuits in January 2026 it is building error-corrected gate-model computers as well, with an initial gate-model system promised for 2026.

History and background

D-Wave was founded in 1999 in Canada as a spin-off from research at the University of British Columbia, and it spent its first decade developing superconducting annealing chips. In 2011 it sold the D-Wave One to Lockheed Martin, which is generally described as the first commercially sold quantum computer. Google, NASA, and other organizations bought or hosted later machines. Skeptics argued for years about whether the devices were truly faster than classical solvers, and the debate has never fully ended.

The company went public in 2022 through a merger with a special purpose acquisition company. Alan Baratz became CEO in 2020 and shifted the business toward cloud access through its Leap service and toward hybrid solvers that combine quantum and classical computing. The current annealing system, Advantage2, became widely available in 2025. In 2026 D-Wave reorganized around a dual-platform strategy by acquiring Quantum Circuits, a Yale-linked startup, for 250 million dollars in cash plus roughly 10.4 million shares.

How the technology works

In annealing, a problem is encoded into the couplings between qubits. The qubits start in a simple quantum state and are slowly steered, using a controlled magnetic field, toward a configuration that represents a low-cost answer. Quantum effects such as tunneling may help the system escape poor local solutions. The output is a sample of good answers rather than a single guaranteed optimum, which suits some problems and not others.

Advantage2 uses a connectivity layout called Zephyr and has more than 4,400 qubits. D-Wave reports that, compared with the previous generation, it has about 40% higher energy scale, roughly twice the coherence time, and four times lower noise. Counting qubits in an annealer is not comparable to counting qubits in a gate-model machine, since annealers have no universal gate set. Advantage2 systems operated in three countries (Canada, the United States, and Germany) at the end of 2025.

The gate-model effort uses dual-rail qubits developed by Quantum Circuits and its co-founder Rob Schoelkopf, now D-Wave's chief scientist. A dual-rail qubit encodes information in which of two resonators holds a photon, which makes the most common errors detectable as erasures and simplifies error correction. D-Wave's roadmap calls for 17 physical qubits in 2026, 49 in 2027, 181 in 2028, and around 10 logical qubits by 2030. It has also described an on-chip control design meant to scale to 100,000 qubits without a proportional increase in wiring.

Products and platforms

Processors in our Hardware Database

• D-Wave Advantage2 — 4,400 qubits, released 2025 (general availability May 20, 2025)

Key milestones

Strengths and challenges

Strengths

• The longest commercial track record of any quantum computing company, with customers in telecom, logistics, and public safety.
• Large cash reserves relative to its revenue, reported at about 546 million dollars at the end of June 2026.
• A new gate-model program with an unusual error-detecting qubit design.
• A cloud service and hybrid solvers that let customers use the technology without owning hardware.

Challenges and open questions

• Annealers cannot run general quantum algorithms, and whether they beat the best classical optimization methods remains debated.
• Revenue is small and lumpy: quarterly revenue was flat at about 3.1 million dollars in Q2 2026 because system sales are timed irregularly.
• The gate-model roadmap starts at only 17 physical qubits in 2026, so it is years behind several competitors.
• Integrating Quantum Circuits and funding two hardware programs will raise costs.

Why it matters

D-Wave is a useful counterexample to the idea that quantum computing is only about gate-model machines. Many real business problems, such as scheduling shifts or routing deliveries, are optimization problems, and annealing is a different tool for them. Whether it actually beats classical solvers on such problems is one of the most important open questions for the commercial side of the field.

The company's move into gate-model hardware also shows how competition is reshaping the industry. A pure annealing business was not enough to satisfy customers or investors who want fault-tolerant machines, so even the oldest quantum company is adding a second platform.

Latest news

• D-Wave reports Q2 2026 results: revenue of 3.1 million dollars, first-half bookings up more than 1,100%, and a design for scaling to 100,000 qubits — August 2026
• D-Wave announces a forthcoming gate-model simulator and a National Science Foundation grant of about 1.57 million dollars — 2026
• D-Wave publishes a peer-reviewed Nature paper on a high-fidelity two-qubit gate in its dual-rail architecture — 2026
• D-Wave completes the acquisition of Quantum Circuits — January 2026

Frequently asked questions

This profile is written for educational purposes and summarizes public announcements, company filings, and published research. Last reviewed: September 30, 2026. It is not investment advice. Check the company's official channels for the latest figures and claims.

← Back to all quantum companies