The Zhitong Finance App learned that the US quantum computing leader D-Wave Quantum (QBTS.US), which focuses on quantum annealing and “quantum classical computation hybrid optimization,” soared 13% before the US market on Monday. Earlier, the company announced a new cooperation agreement with AT&T (T.US), one of the top three telecom operators in the US, to expand the telecom supergiant's use of its quantum computing technology in optimizing high-speed network technology. However, the announcement did not disclose the amount and duration of the contract.
According to information, AT&T initially plans to integrate D-Wave's quantum annealing calculation technology into the tools that support its intelligent AI solutions (that is, Agentic AI created exclusively by AT&T). Improving fault detection and network management techniques through quantum computing technology, these solutions have helped reduce customer service interruptions of approximately 12 million hours in 2025.
The added quantum computing capabilities may help further improve the operating efficiency of AI agent-type proxy workflows in AT&T network operations.
In an early deployment, D-Wave reduced the processing time lifecycle for a network optimization workload from about an hour to less than 15 seconds. AT&T is currently planning to evaluate the company's cutting-edge quantum technology in more application scenarios, including fault detection and instantaneous response, technician route planning, network infrastructure construction planning, and traffic management.
According to information, D-Wave's core track is quantum annealing (quantum annealing) and quantum-classical hybrid optimization, rather than the general-purpose gate model quantum computation that IBM, Google, and Quantinuum are currently betting on.
The company's quantum annealing technology transforms problems such as network scheduling, route planning, resource allocation, and combination selection into the form of an ISing model or QuBO to obtain approximate optimal solutions by searching for the low energy state of the system, which is particularly suitable for combination optimization with many variables and complex constraints. AT&T's technician dispatch, fault response, and network traffic management fall into this category: early testing reduced specific optimization loads from about an hour to less than 15 seconds, so the collaboration expanded from the verification phase to more network operation scenarios. D-Wave is also developing a gate model route, but the most commercially recognizable at this stage are still annealing systems and hybrid solvers.
A series of technological advances and developments from leading quantum computing companies such as Nvidia, Cisco, IBM, Google, Quantinuum, D-Wave, and Pasqal are essentially laying the software and hardware infrastructure for controllable commercial quantum computing systems that may appear in 2030. In other words, quantum computing is becoming the next generation of “grand technology narratives” most sought after the AI superwave.
Quantum computing is developing rapidly, and funding is focused on the “next-generation computational revolution”
Quantum computing — widely regarded by Wall Street analysts as the core engine of the “next-generation computing revolution”. Although it is still in the early stages of development, the accelerated breakthroughs in this cutting-edge technology and the popularity of capital are resonating. The “quantum computing boom” has moved from a scientific research narrative to a new round of technology stock narratives of financing, listing, and valuation expansion.
Global quantum computing is in the “post-NISQ, pre-large-scale fault-tolerant quantum computation” engineering transition period. NISQ is an English abbreviation for “noisy medium-scale quantum,” proposed by physicist John Preskill (John Preskill) in 2018. It is used to describe current and recently owned quantum computers. Its characteristics are that the number of qubits is between tens and thousands, there is noise (error) in operation, and large-scale quantum error correction is not possible.
Current equipment is no longer just a small number of physical qubits in the laboratory; the focus of technology is shifting from “increasing the number of qubits” to manufacturing logical qubits with a low error rate. The Google Willow platform has proven on 72 and 105 physical qubit systems that when the scale of error correction codes is expanded, the logic error rate can be reduced, that is, across the so-called critical threshold of “below the error correction threshold”; experiments by Microsoft and Quantinium reduced the error rate by about 11 times to 800 times compared to physical circuits in several logic circuits. These developments prove that fault-tolerant quantum computation is physically possible, but there is still a huge engineering distance from having hundreds to thousands of high-quality logical qubits and continuously executing deep algorithms.
As a result, there are actually two parallel commercial paths today: D-Wave annealing and hybrid systems are already being deployed in narrow fields such as scheduling, logistics, networks, and resource optimization; the general-purpose gate model system is still mainly in the stages of error correction, quantum simulation, and algorithm verification. IBM plans to launch Starling in 2029, which has about 200 logical qubits and can run 100 million quantum gates, but this is a roadmap rather than a delivered capability.
Academic research at the quantum physics level has yet to prove that quantum annealing can steadily surpass the best classical optimization algorithms on a wide range of practical problems. This is why AT&T's 15-second case has important commercial signal significance, but it cannot be equated alone with a universal “quantum advantage”; it is also necessary to compare the quality of end-to-end solutions, hardware costs, data mapping time, and the most advanced classical algorithms.
McKinsey estimates that global quantum computing companies' revenue in 2025 has exceeded 1 billion US dollars, more than 300 companies are cooperating with quantum technology companies, and related revenue may increase to 4.4 billion US dollars by 2028. This means that quantum as a service, consulting, testing platforms, hybrid optimization, and quantum security have formed a real market; however, fault-tolerant quantum computing, which can generally change productivity in chemistry, pharmaceuticals, finance, and materials science, still depends on logical qubit scale, error correction costs, control systems, low temperature engineering, chip interconnection, and whether algorithms can economically beat classical computation that continues to advance.
The “quantum industry” that actually first achieved large-scale revenue may not even be quantum computing itself, but rather a post-quantum cryptographic migration. NIST has officially released the three major post-quantum cryptography standards ML-KEM, ML-DSA, and SLH-DSA, and has urged institutions to begin migrating; such cybersecurity, cryptographic upgrades, and cryptographic asset inventory requirements do not need to wait for large-scale quantum computers to appear, so their commercial certainty is usually higher than betting on a certain qubit technology route.