How AT&T Uses D-Wave Quantum Annealing and Agentic AI to Optimize Network Operations in Under 15 Seconds
AT&T used D-Wave quantum annealing and agentic AI to cut network optimization from one hour to under 15 seconds.
Quantum Leaves the Lab: AT&T’s Real‑World Network Speedup with D‑Wave
1. Summary
In July 2026, AT&T (NYSE:T) and D‑Wave Quantum Inc. (NASDAQ:QBTS) announced an expanded commercial agreement to deploy D‑Wave’s quantum annealing technology across AT&T’s live network operations, marking one of the most significant documented instances of production quantum computing in a Tier 1 telecommunications environment. An early application reduced a network optimisation workload from approximately one hour to under 15 seconds, representing a 240‑fold acceleration [1]. This performance gain was achieved on a combinatorial optimisation problem (the class of problems for which quantum annealing is specifically designed) and was integrated with AT&T’s existing agentic AI tooling, which in 2025 alone reduced customer downtime by 12 million hours [1].
The expanded scope encompasses outage detection and response, technician routing, network build planning, and traffic management. AT&T is also evaluating D‑Wave’s forthcoming gate‑model systems for quantum security and communications applications [2][5][6]. D‑Wave, which positions itself as the only commercial supplier offering both annealing and gate‑model quantum computing platforms, saw its stock rise approximately 20% on the announcement [12].
The AT&T‑D‑Wave deployment represents a meaningful validation of quantum annealing’s commercial applicability for specific, narrowly scoped optimisation problems within telecommunications networks. However, the evidentiary basis for scalability to larger, more complex problems remains limited, and the financial terms of the agreement were not disclosed. The strategic significance extends beyond the immediate use case: it signals that quantum computing has crossed a threshold from laboratory research to operational deployment, albeit in a specialised and constrained form. For telecommunications operators, the primary strategic implication is the need to develop internal capability to identify and formulate network optimisation problems amenable to quantum annealing, while maintaining realistic expectations about the technology’s current limitations. For investors, the deal provides commercial validation but does not resolve fundamental questions about D‑Wave’s path to profitability, given first‑quarter 2026 revenue of $2.9 million against a market capitalisation of approximately $7.2 billion [8][12].
2. Contextual and Scientific Background
2.1 The Quantum Computing Landscape
Quantum computing encompasses multiple distinct hardware paradigms, the two most commercially relevant being quantum annealing and gate‑model quantum computing. Gate‑model systems, pursued by IBM (NYSE:IBM), Google (NASDAQ:GOOG), Quantinuum, and others, implement universal quantum computation through sequences of quantum logic gates applied to qubits, analogous to classical digital computing. These systems are theoretically capable of executing any quantum algorithm but face significant engineering challenges in scaling qubit counts while maintaining coherence and managing error rates.
Quantum annealing, by contrast, is a specialised analog approach designed to solve combinatorial optimisation problems. The quantum processor is initialised in a superposition state and gradually evolved toward a classical ground state that encodes the solution to the optimisation problem. This trade‑off trades universality for scale: annealing systems can operate with thousands of qubits, whereas contemporary gate‑model systems typically operate with fewer than 200 qubits.
2.2 D‑Wave’s Hardware Platform
D‑Wave’s Advantage2 system, released in May 2025, features over 4,400 superconducting qubits and more than 40,000 couplers interconnected in the Zephyr topology, which provides 20‑way qubit connectivity compared to the 15‑way connectivity of the previous‑generation Pegasus topology [3]. The system incorporates three major technology upgrades relative to its predecessor: 40% higher energy scales, twofold longer coherence time, and fourfold lower noise. Higher energy scales increase the energy separation between high‑quality and low‑quality solutions, driving results closer to optimal; longer coherence time improves the effectiveness of the quantum annealing algorithm; and lower noise reduces imprecision in representing problem weights.
These hardware improvements are significant. D‑Wave’s internal benchmarking shows that the Advantage2 system produces better‑quality solutions using anneal times several orders of magnitude faster than those used on the previous‑generation Advantage system [4]. However, these performance claims are based on D‑Wave’s own benchmarking of 3D‑lattice spin glass problems and have not been independently verified by third parties for the specific network optimisation workloads AT&T is addressing.
2.3 Quantum Annealing and Network Optimisation
Telecommunications network optimisation presents a natural application domain for quantum annealing because many network management problems can be formulated as quadratic unconstrained binary optimisation (QUBO) problems or as Ising models, which are the mathematical formalisms that annealing quantum computers are designed to solve. Routing and wavelength assignment in optical networks, dynamic spectrum allocation, and minimum edge multiway cut problems (which evaluate network resilience) have all been the subject of peer‑reviewed investigation using quantum annealing approaches.
A 2025 paper in the Journal of Optical Communications and Networking developed four QUBO formulations for routing problems relevant to optical transport layers, demonstrating viability for joint routing and wavelength assignment, unicast and multicast trees, and shared risk avoidance using D‑Wave’s hybrid solver [5]. A 2026 paper in IEEE Communications Magazine outlined a methodology for large‑scale network optimisation using quantum annealing and quantum reinforcement learning, while identifying the main challenges that quantum algorithms and hardware must overcome to effectively optimise future networks [6]. A 2026 study from Institut Polytechnique de Paris found that quantum annealing currently offers the most scalable performance for minimum edge multiway cut problems, while photonic and gate‑based approaches remain limited by hardware and simulation depth [7].
These peer‑reviewed investigations establish that the theoretical and algorithmic foundations for applying quantum annealing to telecommunications network optimisation are sound. The AT&T deployment represents an engineering translation of these foundations into an operational context.
3. Key Players and Stakeholders
3.1 AT&T
AT&T is a Tier 1 U.S. telecommunications operator with a converged fibre and 5G network. The company’s engagement with D‑Wave is part of a broader innovation strategy applying quantum computing, AI, automation, advanced analytics, and software‑defined infrastructure to modernise network operations. AT&T’s agentic AI tools, which reduced customer downtime by 12 million hours in 2025, serve as the integration platform for quantum optimisation capabilities. Lucus Haugen, director of Data Science for AT&T’s Chief Data Office, characterised the speed achieved with D‑Wave as “challenging what’s currently possible” [1].
3.2 D‑Wave Quantum Inc
D‑Wave is the world’s first commercial supplier of quantum computers and the only company offering both annealing and gate‑model quantum computing platforms. As of July 16, 2026, the company had a market capitalisation of approximately $6.26 billion, rising to approximately $7.2 billion following the AT&T announcement [8]. First‑quarter 2026 revenue was $2.9 million, though bookings reached a record $33.4 million, up 1,994% year over year [8]. The company closed the first quarter with $588 million in cash and investments. D‑Wave was named a Leader in the IDC MarketScape: Worldwide Quantum Computing 2026 Vendor Assessment.
3.3 Other Telecommunications Operators
AT&T is not alone in exploring quantum technologies. Comcast completed a trial with AMD (NASDAQ:AMD) and Classiq that leveraged quantum software to find independent backup paths for network sites. Deutsche Telekom and Qunnect successfully demonstrated quantum teleportation over an existing fibre network in Berlin. Telefónica Tech partnered with Qilimanjaro Quantum Tech, Multiverse Computing, and Qcentroid to pursue integration between AI and quantum computing. Telus partnered with Photonic to demonstrate quantum teleportation over 30 kilometres of existing network [9]. These initiatives span quantum optimisation, quantum communications, and quantum key distribution, indicating a broad‑based industry interest that extends beyond any single application or vendor.
3.4 Competing Quantum Computing Vendors
IBM maintains the largest installed base of gate‑model quantum systems accessible via cloud services and has partnered with Cisco to develop quantum networking capabilities. IonQ (NASDAQ:IONQ) and Rigetti (NASDAQ:RGTI) are publicly traded pure‑play quantum computing companies; IonQ has pursued quantum networking applications while Rigetti’s focus remains on cloud‑accessible quantum compute. All three vendors compete with D‑Wave in the broader quantum computing market, though their gate‑model architectures address a different problem class than D‑Wave’s annealing systems. The quantum ecosystem also includes cloud service providers such as Amazon (NASDAQ:AMZN) (which offers access to Rigetti, IonQ, and D‑Wave systems through Amazon Braket), software vendors, and academic research groups.
John D
4. Technical and Operational Considerations
4.1 Demonstrated Performance
The central technical claim is the reduction of a network optimisation workload from approximately one hour to under 15 seconds. This 240‑fold acceleration is the most specific performance figure publicly available [1]. Several important qualifications attach to this claim. First, the problem being solved has not been fully specified in public sources; without knowing the size, complexity, and structure of the optimisation problem, it is impossible to assess whether this level of acceleration is likely to generalise. Second, the performance figure represents a single early application, not a systematic benchmark across a range of problem instances. Third, the figure has been reported by D‑Wave and AT&T through a D‑Wave press release; no third‑party verification of the specific 15‑second figure has been published. The report treats this as a demonstrated result from the parties involved, subject to these evidentiary limitations.
4.2 Hybrid Quantum‑Classical Workflow
The AT&T deployment employs a hybrid quantum‑classical workflow in which D‑Wave’s annealing quantum processor handles the optimisation core while classical systems manage data pre‑processing, problem formulation, and solution validation. This hybrid approach is standard in commercial quantum computing deployments because current quantum processors are not standalone general‑purpose computers; they are specialised accelerators for specific computational kernels. AT&T’s integration of D‑Wave’s annealing capabilities into its existing agentic AI tooling suggests that the quantum optimisation step is invoked as a subroutine within a broader AI‑driven operational framework.
4.3 Planned Expansion
AT&T plans to explore quantum annealing across outage detection and response, technician routing, network build planning, and traffic management [1]. These applications span both real‑time operational problems (outage response, traffic management) and longer‑term planning problems (network build planning). The diversity of planned applications suggests that AT&T views quantum annealing as a broadly applicable optimisation tool rather than a solution to a single isolated problem. However, these planned applications remain aspirational at the time of this report; no performance data for these additional use cases has been disclosed.
4.4 Gate‑Model Roadmap and Quantum Security
AT&T is evaluating D‑Wave’s forthcoming gate‑model systems for potential applications in quantum security and quantum communications [2]. D‑Wave’s gate‑model roadmap targets a 17‑physical‑qubit system in 2026 capable of achieving logical error rates two times lower than physical error rates, scaling to a 49‑physical‑qubit array in 2027 (20x error reduction), and a 181‑physical‑qubit architecture in 2028 establishing a 2,000‑fold error suppression blueprint. The long‑term target is 100 logical qubits capable of performing over 1 million operations by 2032 [2]. D‑Wave also plans to launch a gate‑model quantum computing simulator via its Leap cloud platform in September 2026.
These gate‑model developments are distinct from the annealing deployment and should be evaluated separately. The gate‑model roadmap describes future capabilities, not demonstrated current performance. The potential applications in quantum security and cryptographically relevant quantum algorithms depend on the successful execution of this roadmap, which remains subject to substantial technical uncertainty.
4.5 Distinguishing Demonstration from Projection
The AT&T‑D‑Wave deployment demonstrates that quantum annealing can solve a specific network optimisation problem faster than the prior classical approach [1]. It does not demonstrate that quantum annealing outperforms all possible classical approaches, that the observed speedup will scale to larger problems, or that quantum annealing provides a general advantage across the full range of network optimisation tasks. The planned expansion into additional applications represents an exploratory programme, not a validated capability. The gate‑model security applications are aspirational and contingent on successful hardware development [2].
5. Economic and Market Dynamics
5.1 Market Reaction
The company’s market capitalisation reached approximately $7.2 billion following the announcement [8]. Wall Street maintained a consensus “Strong Buy” rating on QBTS stock heading into the announcement, with a mean price target of nearly $37 [12]. However, as of the announcement date, D‑Wave stock had retreated over 35% in 2026, indicating significant volatility and suggesting that the market has not uniformly embraced the commercial quantum computing thesis.
5.2 Financial Position and Revenue Trajectory
D‑Wave’s first‑quarter 2026 revenue was $2.9 million, declining from the prior year due to a $12.6 million one‑time quantum computer sale in the previous quarter. Bookings reached a record $33.4 million, up 1,994% year over year, including a $20 million system purchase by Florida Atlantic University and a $10 million two‑year Quantum Computing as a Service agreement with a Fortune 100 company [8]. The company reported a net loss of $18.4 million in the first quarter as operating expenses increased following the acquisition of Quantum Circuits. With $588.4 million in cash and investments, D‑Wave has substantial runway.
The AT&T deal’s financial terms were not disclosed [1]. This absence of financial specificity limits the ability to assess the deal’s materiality to D‑Wave’s revenue. The deal’s primary value to D‑Wave may be commercial validation and reference‑ability rather than immediate revenue contribution.
5.3 Market Size Estimates
The global quantum computing hardware and services market is estimated at approximately $1.42 billion in 2024, with projections reaching $8.5 billion by 2030, representing a compound annual growth rate of 34.8% [10]. These figures come from third‑party industry analysts and should be treated as projections rather than established facts. The telecommunications sector represents a subset of this market, with specific estimates varying widely across sources. The AT&T‑D‑Wave deployment provides a concrete data point supporting the thesis that quantum computing can generate operational value in telecommunications, but it does not by itself validate market size projections.
6. Regulatory Landscape
The primary regulatory consideration relevant to the AT&T‑D‑Wave deployment is the standardisation and mandated migration to post‑quantum cryptography (PQC). The National Institute of Standards and Technology (NIST) finalised the first three post‑quantum cryptographic standards in August 2024. In May 2026, NIST selected nine candidates for the third round of the additional digital signatures standardisation process, with this evaluation phase expected to last approximately two years [11].
The U.S. government has established hard migration deadlines: December 31, 2030, for key establishment and December 31, 2031, for digital signatures [11]. The Federal Acquisition Regulation Council has 180 days to publish a proposed rule requiring covered contractors to comply with NIST PQC‑related FIPS standards by December 31, 2030. NIST will run a PQC migration pilot to be completed by December 31, 2027 [11].
Industry risk assessments generally converge on a threat window around 2030, when quantum computers may become capable of breaking existing public‑key cryptography [10]. The Alliance for Telecommunications Industry Solutions has examined quantum computing technologies and found that successful development could accelerate the threat timeline.
For the specific AT&T‑D‑Wave deployment of annealing quantum computing for network optimisation, no sector‑specific telecommunications regulation directly governs the use of quantum annealing. The regulatory relevance is indirect: AT&T’s evaluation of D‑Wave’s gate‑model systems for quantum security applications aligns with the broader industry imperative to prepare for post‑quantum cryptographic migration. However, the annealing deployment itself does not implicate cryptographic functions and therefore does not directly engage PQC regulatory requirements.
7. Geopolitical and Strategic Dimensions
The strategic significance of quantum computing for national competitiveness and security is substantial, though the AT&T‑D‑Wave deployment’s direct geopolitical implications are limited. The quantum landscape is irreducibly multipolar: the United States leads in quantum computing platforms; China leads in quantum communications deployment; Europe and Japan anchor critical component supply chains in cryogenics, lasers, optics, and detectors [10]. No single nation can control the full quantum technology stack.
U.S. public investment in quantum technologies has been substantial, anchored by the National Quantum Initiative Act. China has integrated quantum information into national five‑year plans as a major priority, with achievements including the Micius satellite, the Beijing‑Shanghai backbone network, and the Jiuzhang and Zuchongzhi series of quantum supremacy experiments. The European Union is implementing a ten‑year Quantum Flagship programme and is expected to introduce a European Quantum Act in 2026 [10]. EU public investment in quantum technologies as of 2024 trailed only China’s, at $15 billion versus higher Chinese figures.
D‑Wave’s position as a U.S.‑based quantum computing company with both annealing and gate‑model capabilities contributes to U.S. competitiveness in the quantum computing segment. The AT&T deployment demonstrates a U.S. commercial application of quantum technology in a critical infrastructure sector. However, the annealing paradigm in which D‑Wave specialises addresses a narrower problem class than the universal quantum computing that underpins most strategic quantum competition narratives. The geopolitical significance of this specific deployment should not be overstated; it is one data point in a broader and more complex competitive landscape.
8. Risk Matrix
| Risk Category | Risk Description | Likelihood | Impact | Mitigations |
|---|---|---|---|---|
| Technical: Scaling | Quantum advantage may not scale to larger, more complex network optimisation problems | Medium‑High | High | Phased deployment with clear success criteria; maintain classical fallback options; rigorous benchmarking across problem sizes |
| Technical: Integration | Integration with existing agentic AI and operational systems may encounter unforeseen complications | Medium | Medium | Hybrid architecture with clear interfaces; incremental integration; dedicated engineering resources |
| Technical: Gate‑Model Execution | D‑Wave’s gate‑model roadmap may fail to deliver on performance or timeline targets | Medium‑High | Medium (for AT&T) / High (for D‑Wave) | Treat gate‑model as exploratory; maintain alternative quantum security approaches; avoid lock‑in |
| Commercial: Revenue Impact | Limited revenue impact from AT&T deal; financial terms undisclosed | High | Medium (for D‑Wave) | Diversify customer base; focus on bookings growth; manage investor expectations |
| Commercial: Customer Concentration | Over‑reliance on a small number of large customers | Medium | Medium | Expand enterprise sales; develop vertical‑specific solutions; leverage cloud distribution |
| Competitive: Classical Alternatives | Classical optimisation algorithms or specialised hardware may match or exceed quantum performance | Medium | Medium | Continuous benchmarking against classical baselines; focus on problems where quantum provides unique advantage |
| Competitive: Rival Quantum Vendors | IBM, IonQ, Rigetti, or others may develop superior annealing or gate‑model solutions | Medium | Medium | Maintain technology differentiation; invest in R&D; build switching costs through integration |
| Security: Quantum Threat to Encryption | Cryptographic vulnerability window (circa 2030) may arrive sooner than expected | Low‑Medium | High | Implement PQC migration planning; monitor quantum threat assessments; prioritise crypto‑agility |
| Security: Quantum‑Safe Migration | Failure to migrate to PQC by regulatory deadlines | Low | High | Develop PQC roadmap; allocate resources; engage with standards bodies |
9. Strategic Recommendations
9.1 For Telecommunications Operators and Industrial Users
Telecommunications operators should evaluate quantum annealing for network optimisation through a structured, evidence‑based process. The AT&T deployment demonstrates that quantum annealing can deliver meaningful speedups for specific combinatorial optimisation problems, but it does not establish a general case for quantum advantage. Operators should:
Identify and formulate suitable problems. Network optimisation problems that can be expressed as QUBO or Ising models (including routing, wavelength assignment, spectrum allocation, and technician scheduling) are the most natural candidates. Operators should conduct internal audits of their optimisation workloads to identify problems with high computational cost and clear QUBO formulations.
Build hybrid quantum‑classical competence. The AT&T model of integrating quantum optimisation into existing agentic AI tooling is instructive. Operators should invest in the engineering capability to integrate quantum solvers into operational workflows rather than treating quantum as a standalone capability.
Maintain realistic expectations and classical baselines. The observed 240‑fold acceleration is impressive but represents a single data point. Operators should benchmark quantum performance against the best available classical approaches, including specialised classical solvers and heuristic algorithms, and should not assume that quantum will outperform across all problem instances or scales.
Engage with multiple quantum vendors. While D‑Wave is the leading annealing vendor, operators should maintain awareness of gate‑model developments from IBM, IonQ, and others, as well as quantum‑inspired classical algorithms. Technology lock‑in is premature in this rapidly evolving field.
Plan for post‑quantum cryptography migration. The 2030‑2031 regulatory deadlines are approaching. Operators should develop PQC migration roadmaps, inventory cryptographic assets, and begin testing PQC algorithms in non‑production environments.
9.2 For Investors and Corporate Strategists
The AT&T‑D‑Wave deal provides commercial validation for quantum annealing but does not resolve the fundamental investment questions surrounding D‑Wave and the quantum computing sector more broadly. Investors should:
Distinguish between annealing and gate‑model quantum computing. D‑Wave’s annealing business has demonstrable commercial traction, as evidenced by the AT&T deal and record bookings. The gate‑model business is at an earlier stage and carries substantially higher technical risk. These are distinct investment theses within the same company.
Evaluate the revenue‑to‑valuation ratio critically. D‑Wave’s market capitalisation of approximately $7.2 billion against first‑quarter revenue of $2.9 million implies extraordinary growth expectations [8][12]. The AT&T deal provides validation but does not, in itself, justify the valuation. Investors should monitor revenue conversion from bookings, customer diversification, and gross margin trends.
Assess the competitive moat. D‑Wave’s first‑mover advantage in annealing quantum computing is significant, but the technology is not immune to competition from classical alternatives or from other quantum vendors. The company’s dual‑platform strategy (annealing plus gate‑model) is a differentiator, but execution risk is substantial.
Monitor the gate‑model roadmap closely. The 2026 delivery of a 17‑physical‑qubit system with logical error rates two times lower than physical error rates is a critical milestone [2]. Success would validate D‑Wave’s gate‑model approach; failure would reinforce the view that D‑Wave is primarily an annealing company. The September 2026 gate‑model simulator launch is an earlier indicator.
Consider the broader quantum computing market dynamics. The quantum computing sector is characterised by high technical uncertainty, long development timelines, and significant capital requirements. D‑Wave’s $588 million cash position provides runway, but the company will need to demonstrate a clear path to profitability. The AT&T deal is a positive data point but not a turning point.
Watch for regulatory and geopolitical catalysts. PQC mandates, national quantum strategies, and export controls could affect market dynamics. D‑Wave’s U.S. base and defence‑related applications (for example, the Advantage2 deployment at Davidson Technologies for U.S. defence applications) position it favourably in the context of U.S. strategic competition, but geopolitical tensions could also disrupt supply chains or market access.
The Quant
John D
References
[1] D‑Wave Quantum Inc. 2026. “AT&T Signs Agreement to Expand Use of D‑Wave’s Quantum Computing Technology Across Network Operations.” Press release. July 27, 2026.
[2] D‑Wave Quantum Inc. 2026. “D‑Wave Charts a New Course to Fault‑Tolerant Quantum Computing with Gate‑Model Roadmap.” Press release. June 1, 2026.
[3] D‑Wave Quantum Inc. n.d. “D‑Wave’s Advantage2 Quantum Computer Now Generally Available.” Support documentation. Accessed July 2026.
[4] D‑Wave Quantum Inc. 2025. “Performance Gains in the D‑Wave Advantage2 System at the 4,400‑Qubit Scale.” Whitepaper. May 12, 2025.
[5] Davies, Ethan, Darren Banfield, Ben Weaver, Catherine White, and Nigel Walker. 2025. “Routing and Wavelength Assignment Problems in Optical Networks—Comparing Formulations for Solution by Quantum Annealing.” Journal of Optical Communications and Networking 17 (12): B83‑B91.
[6] IEEE Communications Magazine. 2026. “Quantum Computing for Large‑Scale Network Optimization: Opportunities and Challenges.” 64 (1): 116‑122. January 2026.
[7] Institute Polytechnique de Paris. 2026. “Quantum Approaches to the Minimum Edge Multiway Cut Problem.” Research portal. January 1, 2026.
[8] Macrotrends. 2026. “D‑Wave Quantum Market Cap 2021‑2026.” July 16, 2026.
[9] Fierce Network. 2026. “Quantum Telecom: What’s New from Comcast, Deutsche Telekom and Qunnect.” February 20, 2026; and “AT&T Joins Growing List of Telcos Making Quantum Moves in 2026.” July 27, 2026.
[10] Hoover Institution. 2026. “The Quantum Revolution: A Guide for Allied Policymakers.” July 7, 2026.
[11] National Institute of Standards and Technology. 2026. “Nine Candidates Advance to the Third Round of the Additional Digital Signatures for the PQC Standardization Process.” May 13, 2026; and “Status Report on the Second Round of the Additional Digital Signature Schemes for the NIST Post‑Quantum Cryptography Standardization Process.” May 14, 2026.
[12] Mitrade. 2026. “D‑Wave Quantum Stock Forecast: Can the AT&T Deal Justify QBTS’ $7 Billion Valuation?” July 28, 2026.
