AT&T’s 240X Quantum Speedup Revives an Old D-Wave Fight

AT&T cut an hour-long network optimization job down to under 15 seconds using D-Wave’s quantum annealing computer, a jump the two companies are calling a 240-fold speedup. D-Wave announced the results on July 27, alongside a new agreement to expand that technology across AT&T’s live network operations. The news landed on the same day D-Wave began trading on the Nasdaq under its existing ticker, QBTS, after moving its listing from the New York Stock Exchange.

The timing is awkward in one specific way. Weeks earlier, D-Wave was publicly defending its last big “beyond-classical” claim, a peer-reviewed quantum simulation result, against physicists who say a classical algorithm now reproduces it. That is not a new problem for D-Wave. It is the same argument the company has been having, in different forms, for more than a decade.

AT&T Turns an Hour of Optimization Into 15 Seconds

AT&T already runs AI agents that diagnose and fix network problems on their own. Those agentic tools cut customer downtime by 12 million hours in 2025 through better outage detection and management, according to details in the companies’ announcement. D-Wave’s annealing hardware is now being layered directly into that same stack, aimed at the optimization math underneath it.

Quantum annealing is a narrower, analog approach to quantum computing built specifically to search enormous solution spaces for the single best answer, rather than running arbitrary step-by-step programs the way gate-model machines do. AT&T is applying it to problems that shift by the minute, where a classical answer that takes an hour to compute can already be stale by the time it arrives. The company’s plans include:

• Network build planning – deciding where and how to expand infrastructure
• Traffic routing and load balancing – shifting demand across the live network in real time
• Outage rerouting – recalculating paths around failures as they happen
• Spectrum and resource allocation – assigning limited capacity across competing demands
• Technician routing – sequencing field repair work more efficiently

D-Wave chief executive Alan Baratz framed the deal in a statement as evidence of a broader shift, saying AT&T “understands where classical computing is challenged, and is moving quickly to explore where quantum can make an impact.” AT&T’s side made a similar case in blunter terms.

The speed we’re seeing with D-Wave challenges what’s currently possible.

That line came from Lucus Haugen, quoted on AT&T’s behalf in the announcement, who added that the technology has the potential to help the carrier optimize faster and scale more real-time operations as it modernizes its network.

A Nasdaq Debut Turns Into a Rally for Quantum Stocks

D-Wave’s move to Nasdaq had been in the works since mid-July. The company disclosed the voluntary transfer on July 14, closed out trading on the NYSE on July 24, and opened on Nasdaq on July 27 under the unchanged QBTS ticker. The AT&T news landed the same morning, and the two events fed each other.

Shares closed the day up more than 20%, on trading volume roughly 54% above their three-month average. The rally spread to other quantum computing stocks that had nothing to do with the AT&T contract.

D-Wave itself has now gained roughly 92% since its 2020 public debut, with shares still ranging widely between $12.75 and $46.75 over the past year. The reception contrasts with what greeted Quantinuum’s own Nasdaq arrival, which opened hot before closing flat instead of building through the session.

Hasn’t Quantum Computing Heard This Before?

Yes, repeatedly. Quantum computing’s biggest speed claims tend to follow the same arc: a company announces a result classical computers supposedly cannot match, then a rival lab or research group narrows or erases the gap once it looks closely. D-Wave has been at the center of that argument since long before AT&T ever signed on.

1. 2019: Google says its Sycamore processor completed a sampling task in 200 seconds that would take a classical supercomputer 10,000 years; IBM disputes the estimate within days, arguing its own Summit machine could do it in about 2.5 days.
2. 2024: A team running the same task on a cluster of 1,432 GPUs beats Sycamore’s original hardware, a result researchers called the first strong refutation of Google’s claim.
3. 2025: D-Wave publishes a peer-reviewed Science paper, “Beyond-classical computation in quantum simulation,” arguing its Advantage2 annealer produced results no classical computer could feasibly reproduce.
4. May 2026: D-Wave issues a public statement defending that paper after researchers publish new classical simulation work challenging it.
5. July 27, 2026: D-Wave and AT&T announce the 240x optimization speedup, the same week D-Wave’s stock moves to Nasdaq.

Long before any of that, theoretical computer scientist Scott Aaronson was already picking apart D-Wave’s early hardware claims. In a 2013 documentary note on the company’s technology, he said the evidence suggested a quantum annealer built without proven speed gains was entirely possible, calling that outcome academically interesting on its own even without a performance edge.

Physicists Are Still Fighting Over D-Wave’s Last Big Claim

The current round of the argument is not settled. It centers on whether a classical algorithm built by researchers connected to the Flatiron Institute can match the physics D-Wave says only its quantum annealer could produce.

• The Flatiron Institute side: a classical method combining 3D tensor networks with a revived 1982 belief-propagation technique reproduces the spin-glass dynamics D-Wave’s Advantage2 annealer generated in its 2025 Science paper.
• D-Wave’s side: Baratz says that work is a meaningful classical advance but does not reproduce the full scope of problem classes and hardest cases the original paper covered, calling claims that the result was overturned “inaccurate.”

D-Wave’s own figures give a sense of the gap it says still exists. The company has said that matching its annealer’s output using classical matrix product state methods, for the largest problems in the study, would take nearly a million years on the Frontier supercomputer. Whether that framing holds up is exactly the kind of question that took years to resolve after Google’s 2019 claim, and it remains unresolved now. Aaronson’s own record on the subject stretches back further still, including a decade of blogged skepticism toward D-Wave that predates this specific dispute by years.

AT&T Hasn’t Published a Classical Benchmark Either

The AT&T claim is a different kind of test than a physics paper. Nobody has to reproduce a peer-reviewed result to judge it. AT&T is a paying customer that chose to expand a contract after seeing internal results, which is a real signal on its own.

But none of the materials released alongside the announcement, not AT&T’s statement, not D-Wave’s, include a published classical baseline showing how a well-tuned classical solver performs on the same specific network optimization problem. That is precisely the gap that fueled every prior dispute on this list, from IBM’s response to Google in 2019 to the Flatiron Institute’s challenge this year. A 240x improvement over an unspecified classical method is a different claim than a 240x improvement over the best classical method available.

AT&T is also evaluating D-Wave’s forthcoming gate-model machines for quantum security and communications, a longer-term bet that points toward the same technology now being watched for its eventual ability to break the encryption protecting systems like Bitcoin wallets, a threat researchers say is compressing from decades away to a nearer-term concern. That evaluation is still exploratory, with no timeline attached.

IonQ and Rigetti Ride D-Wave’s Coattails

D-Wave says more than 100 organizations across commercial, government and research now use its systems, and it remains the only vendor selling both annealing and gate-model quantum platforms. That dual-track position is part of why competitors’ stocks moved on someone else’s contract announcement.

IonQ and Rigetti Computing build gate-model machines, not annealers, and neither company is named in the AT&T deal. Their shares rose anyway, on the assumption that one carrier’s willingness to pay for quantum hardware makes the next enterprise sale, wherever it lands, more likely. That is a real halo effect, but it is also a reminder that quantum stocks currently trade as a group as much as on individual contracts.

D-Wave’s Next Test Comes August 6

D-Wave is scheduled to report second-quarter 2026 earnings on August 6, with analysts expecting a narrower per-share loss even as revenue recognition stays uneven deal to deal. That report will be the first chance to see whether Wall Street’s reaction to the AT&T news and the Nasdaq move shows up in actual bookings.

The bigger test runs longer than one earnings call. D-Wave’s history shows that its boldest claims eventually get checked, sometimes by a rival lab, sometimes by a cluster of GPUs, sometimes by a physicist with a blog. AT&T’s 15-second number is now part of that same record, waiting on its own classical rebuttal or its own confirmation.

Frequently Asked Questions

What is quantum annealing, and how is it different from IBM’s or Google’s quantum computers?

Quantum annealing is an analog approach built to search huge numbers of possible answers and settle on the lowest-energy, best-fit solution, which makes it suited to optimization problems like routing or scheduling. Gate-model machines, the kind IBM, Google, IonQ and Rigetti build, are universal digital systems that execute step-by-step logical operations and can in theory run any quantum algorithm, not just optimization.

Does AT&T’s 240x claim count as scientific quantum advantage?

Not in the strict, peer-reviewed sense that term usually carries. It is a business result reported by the two companies involved, without an accompanying published comparison against a tuned classical solver, which is the kind of adversarial testing that has decided every prior quantum supremacy dispute.

Is D-Wave profitable yet?

No. Analysts covering the company’s upcoming August 6 earnings report are expecting a narrower per-share loss rather than a profit, with revenue that has been described as lumpy from deal to deal.

Are other telecom companies experimenting with quantum computing?

AT&T’s expanded deal follows a wider pattern of telecom operators piloting quantum tools in 2026, though most of that activity remains in early testing rather than full production deployment of the kind AT&T is now describing.

Disclaimer: This article is for informational purposes only and is not investment advice. Quantum computing stocks, including D-Wave, IonQ and Rigetti, are volatile and speculative; consult a licensed financial adviser before making investment decisions. Figures are accurate as of publication on July 28, 2026.