How Powerful Is Quantum Computing, Really?

The history of computing suggests that transformative technologies rarely arrive fully formed. In 1947, the first transistor looked like a laboratory curiosity. Today, more than 100 sextillion transistors power the world’s digital economy. Quantum computing may follow a different path, but governments are not waiting to find out. The United States, China and Europe have already committed well over $70 billion to quantum research and commercialization. NIST wants vulnerable encryption largely phased out by 2035, while researchers warn that a cryptographically relevant quantum computer could emerge years sooner than many organizations are prepared for. Will quantum become so powerful at the handful of problems they can solve that they reshape finance, medicine, cybersecurity and warfare?

On April 29, 2026, Scott Aaronson received the kind of honor that arrives once in a career. The University of Texas at Austin computer scientist was elected to the National Academy of Sciences, one of the highest distinctions in American research. For two decades he had built his reputation on a single, stubborn habit: telling investors, journalists and fellow physicists that quantum computers could not yet do most of what Silicon Valley and Wall Street claimed. His blog, Shtetl-Optimized, became the place serious researchers went to get a quantum claim stress tested by the field’s most reliable skeptic.

Then, less than a day after his election was announced, Aaronson published a post that undercut his own decades of caution. People whose judgment he trusted more than his own on quantum hardware, he wrote, now believed a fault tolerant quantum computer capable of breaking the encryption protecting most of the internet ought to be possible by around 2029. He signed off warning readers not to say later that he had failed to warn them.

It was a strange moment for the field. The world’s most credentialed skeptic had just told everyone to stop waiting for proof and start preparing for consequences. And it captured, in miniature, the confusion sitting at the center of quantum computing in the summer of 2026: an industry that produces astonishing laboratory results and almost no commercial revenue, that four national governments now treat as a matter of strategic survival, and that most people still cannot describe in a single accurate sentence.

So, how powerful is quantum computing, really?

Not a Faster Computer, a Different Kind of Machine

The honest answer starts with what a quantum computer is not. It is not a faster version of the laptop on your desk. Classical computers store information as bits, each one a 0 or a 1. Quantum computers use qubits, which can hold a combination of both states at once, and which can be linked together through a property called entanglement so that measuring one instantly affects the others. That combination lets a quantum computer explore an enormous number of possibilities simultaneously, but only for certain kinds of problems: factoring very large numbers, simulating the behavior of molecules, searching unsorted data, and sampling from probability distributions that are effectively impossible for classical machines to replicate.

For everything else, ordering a coffee, running a spreadsheet, rendering a video, classical computers remain faster, cheaper and more reliable, and will stay that way indefinitely. Quantum computers are not a replacement for classical computing. They are a specialized tool aimed at a narrow set of problems where the physics itself provides shortcuts no classical algorithm can match. Whether that narrow set turns out to be economically important is the entire debate.

One useful comparison, offered by researchers trying to explain the field to non-specialists, is to imagine searching a maze. A classical computer checks one path at a time, backtracking whenever it hits a dead end. A quantum computer, in a sense, explores many paths at once, with the mathematics of quantum mechanics arranged so that wrong answers cancel each other out while the right answer reinforces itself, a process called interference. That trick only works for mazes with the right mathematical shape. Most everyday computing problems do not have that shape, which is why quantum computers will never replace the phone in your pocket, only supplement it for a specific, still-narrow slice of scientific and cryptographic problems.

Four Giants, Four Different Bets

The four companies most often invoked in quantum marketing, Microsoft, Google, IBM and Amazon, are not actually pursuing the same technology, which is itself a sign of how unsettled the underlying physics remains. Google and IBM have both bet on superconducting circuits, chips cooled to near absolute zero and etched with loops of resistance-free metal, and have scaled that approach furthest: IBM’s Condor processor packed 1,121 qubits onto a single chip, while Google’s Willow line has pushed past 1,000. IonQ and Quantinuum have instead bet on trapped ions, individual charged atoms suspended and manipulated with lasers, an approach that sacrifices some scaling speed for higher fidelity per qubit.

Microsoft has made the most contrarian wager of all. Rather than scale up superconducting circuits, the company is pursuing topological qubits built from an exotic quasiparticle called a Majorana zero mode, which theoretically protects quantum information by spreading it across the physical structure of the material itself rather than storing it in one fragile location. Microsoft unveiled its prototype, Majorana 1, in February 2025, describing it as designed to scale to a million qubits on a single chip. Independent physicists have been notably more cautious than Microsoft’s own press materials, since Majorana based qubits remain unproven at any commercially meaningful scale, but if the approach works as theorized, it could sidestep much of the error correction overhead burdening its superconducting rivals.

Amazon has taken a different kind of bet entirely, choosing not to pick a hardware winner at all. Its Braket service, run through Amazon Web Services, functions as a marketplace giving customers cloud access to hardware built by IonQ, Rigetti, QuEra and others, alongside AWS’s own in-house research into next-generation error correction. It is a hedge rather than a wager, a bet that whoever eventually builds the winning architecture, Amazon will already be the store shelf it ships from. Microsoft has adopted a similar hybrid posture with Azure Quantum, reselling access to IonQ, Quantinuum and Atom Computing hardware even as it pursues its own long-shot topological chip in parallel.

What the Hardware Can Actually Do Today

The most credible recent claim came from Google, whose Willow chip demonstrated what physicists call below threshold error correction in December 2024, meaning that adding more physical qubits reduced the overall error rate instead of making it worse, a result published in Nature and treated across the field as the clearest sign yet that a large scale, error corrected quantum computer is physically achievable rather than merely theoretical.

Google followed that with a benchmark called Quantum Echoes, which it says ran roughly 13,000 times faster than the best available classical estimate on a specific molecular measurement task, a result the company calls a verifiable quantum advantage because other quantum machines can reproduce it. IBM has disputed similar classical comparisons in the past, arguing that some of Google’s headline benchmarks could be simulated on classical supercomputers far faster than advertised once researchers were motivated to try. IBM has staked its own roadmap on a 120 qubit processor called Nighthawk, which it says is designed to demonstrate quantum advantage on a practical computation by the end of 2026.

China has kept pace on the same narrow measure. Zuchongzhi 3.0, built by the University of Science and Technology of China, posted a 2025 result that researchers describe as competitive with Google’s Willow, while the Jiuzhang photonic line has repeatedly demonstrated sampling problems that would take classical supercomputers billions of years to replicate. Neither achievement corresponds to a task anyone actually needs solved. What they show, researchers say, is that quantum hardware can now operate in a regime where classical simulation becomes genuinely infeasible, which is the precondition for everything else the field hopes to build.

The caveat, repeated by nearly every serious physicist in the field, is that these benchmarks were chosen specifically to showcase quantum advantage and bear little resemblance to commercially useful work. A widely cited technical analysis notes that a distance seven surface code, one of the standard error correcting schemes, needs 49 physical qubits to produce a single reliable logical qubit, while breaking a 2048 bit RSA encryption key with Shor’s algorithm is estimated to require somewhere around 1,399 logical qubits under optimized conditions. Multiply those numbers and the gap between today’s laboratory chips and a code breaking machine is measured in the hundreds of thousands of physical qubits, a gap that has narrowed considerably but has not closed.

Q-Day and the Race Nobody Can See Coming

That gap is exactly what worries the people responsible for protecting encrypted information over the long term. Cryptographers use the term “Q-Day” to describe the moment a sufficiently powerful quantum computer first runs Shor’s algorithm against the public key systems, RSA, Diffie-Hellman and their elliptic curve variants, that secure most of the internet, from bank transfers to diplomatic cables. The threat is not only forward looking. Intelligence agencies and criminal groups are already engaged in what security researchers call harvest now, decrypt later, quietly collecting encrypted traffic today with the expectation of unlocking it once a capable quantum computer exists.

The National Institute of Standards and Technology finalized its first three post quantum encryption standards in August 2024, giving the world a concrete, vetted set of quantum resistant algorithms to migrate toward. Dustin Moody, the NIST mathematician who leads the post quantum cryptography project, has urged organizations to start integrating the new standards immediately, warning in a later interview that the migration will be neither quick nor simple.

“It’s gonna be hard, but we gotta get it done.”
Dustin Moody, NIST mathematician and head of the Post-Quantum Cryptography project

Banks have particular reason to move early, since financial records, mortgage contracts, trade secrets and account credentials often need to stay confidential for decades, well past any reasonable estimate of when a cryptographically relevant quantum computer might arrive. Regulators on both sides of the Atlantic have taken notice: the U.S. Treasury Department and the Bank of England have jointly urged banks, insurers and exchanges to accelerate their post quantum preparations, treating the migration less as a physics bet than as a long-duration data retention problem that has to be solved regardless of exactly when Q-Day lands.

NIST’s official planning horizon puts full deprecation of vulnerable algorithms at 2035, with banks, defense contractors and government archives expected to move sooner. Yet adoption is lagging badly: only about 13 percent of organizations have moved post quantum cryptography into production, and roughly 60 percent have not begun a meaningful migration at all, even as the Cybersecurity and Infrastructure Security Agency reported that only 38 percent of Fortune 500 companies had completed even a partial cryptographic inventory by early 2026, up from 12 percent in late 2024.

Against that backdrop, Aaronson’s 2029 warning landed hard, but it has not moved the official goalposts. A detailed review of the field found that NIST, the NSA and Britain’s National Cyber Security Centre have not revised their 2033 to 2035 central estimate upward in response to Willow, Nighthawk or Aaronson’s post. What has changed, the analysis argues, is the credibility of the estimate, not its position: it is no longer scientifically defensible to argue that scalable error correction might simply be impossible. The pathway is now credible. The pathway is still long.

Sensors, Not Just Simulators, Are Already on the Battlefield

While code breaking quantum computers remain years away, a quieter branch of the same physics is already being deployed. In June 2026, the White House signed executive orders directing the Pentagon to field three new types of quantum sensors by 2028 and to help the Energy Department build a quantum supercomputer. The most immediate use case is navigation without GPS, a capability the military considers urgent given how easily satellite signals can be jammed or spoofed in contested environments.

Under a DARPA program called Robust Quantum Sensors, the company Q-CTRL has tested a system called Ironstone Opal that uses atom interferometry, tracking motion by measuring laser pulses split and recombined across clouds of ultracold atoms, a technique that needs no external signal and cannot be jammed or spoofed. In flight testing, the system achieved up to 111 times greater positioning accuracy than a high end classical inertial navigation system when GPS was unavailable, not a laboratory projection but a demonstrated result in the air.

Dr. Robert Compton, a senior technical fellow at Safran Federal Systems, one of the contractors working on the program, described the effort as part of a broader shift underway across the military.

“a critical step forward in delivering operational quantum sensing capabilities.”
Dr. Robert Compton, Safran Federal Systems, on DARPA’s Robust Quantum Sensors program

The urgency is not abstract. “If we’re relying on space-based, GPS-based PNT, then we may be in trouble,” Adm. Christopher W. Grady, then vice chairman of the Joint Chiefs of Staff, told defense industry executives at a National Defense Industrial Association conference, referring to position, navigation and timing systems. Quantum sensing, in other words, is reshaping how militaries operate years before a quantum computer capable of breaking an adversary’s codes is likely to exist.

Astronomical Valuations, Modest Revenue

Venture investors, who priced early rounds in quantum startups years before any of this became a public market story, tend to describe the current mania with a shrug rather than alarm. Their argument, echoed across the industry, is that every deep-tech wave, from the internet to artificial intelligence, has required investors to fund infrastructure years ahead of proven demand, and that a founder or company chasing capital has little incentive to undersell a technology’s promise while that capital is still available. The unusual part of the quantum story is not that hype exists, but that it has migrated so quickly from venture pitch decks into public markets where ordinary retail investors, rather than specialized funds prepared to wait a decade for a return, are now absorbing the volatility.

If the science is a genuine, if narrow, story of progress, the market’s reaction to it has been something else entirely. Pure play quantum companies have posted some of the largest percentage gains of any sector on Wall Street since 2024, even as their underlying businesses remain tiny by any conventional measure. IonQ, the largest of the group by revenue, carries a market capitalization above $21 billion against roughly $130 million to $187 million in trailing revenue, depending on the period measured, and posted an adjusted operating loss well north of $500 million.

THE VALUATION GAP
IonQ (IONQ)	~$21B+ market cap; ~$130-187M trailing revenue; ~303x price-to-sales
Rigetti Computing (RGTI)	~$7-8.5B market cap; ~$7-12.7M revenue, declining year over year
D-Wave Quantum (QBTS)	~$10-12.6B market cap; ~$24-25M revenue, up 179% year over year
Quantum Computing Inc. (QUBT)	~$3B market cap; ~$550,000 trailing revenue

Analysts at the Motley Fool have argued the sector’s price to sales ratios, in the hundreds for some names, assume a level of commercial success that is still a decade or more away, and warned that Rigetti in particular looks stretched given shrinking sales and continued share dilution. The volatility has been extreme in both directions: by mid-July 2026, the same pure play stocks that had surged 300 to 600 percent in a matter of months were sitting 60 to 76 percent below their 52 week highs, a rapid deflation that traders described as a classic hype cycle unwinding into a broader risk off rotation in technology stocks.

Yet the comparisons to prior speculative bubbles come with an important asterisk. Unlike many dot-com era companies, quantum firms are shipping working machines today, even if their customers are mostly government laboratories, universities and a handful of early corporate research partners rather than mass market buyers. D-Wave’s quantum annealing systems, suited to optimization problems rather than general computation, have produced a small but real commercial book, including a recent multimillion dollar system sale to Florida Atlantic University.

Washington Becomes a Shareholder

In May 2026, the U.S. Commerce Department did something the federal government rarely does outside of a financial crisis: it became an equity owner in a speculative technology sector. The department announced letters of intent worth $2.013 billion in incentives under the CHIPS and Science Act, spread across nine companies including IBM, GlobalFoundries, Rigetti, D-Wave, Infleqtion, Quantinuum, PsiQuantum and Diraq, taking a minority, non-controlling equity stake in each one, explicitly framed as a way to enhance returns for American taxpayers.

IBM received the largest allocation, roughly $1 billion in federal funding matched by another $1 billion from the company itself, to build what it calls the nation’s first purpose built quantum foundry. Markets reacted instantly: D-Wave jumped 33 percent, Rigetti 30 percent and Infleqtion 31 percent within two trading sessions of the announcement. One financial commentator noted that for most of American history, Washington has been a customer of innovation, buying fighter jets and rockets, rather than an owner of it, making the quantum stakes a notable departure in industrial policy.

The ripple effects crossed the Pacific almost immediately. Shares of Chinese quantum firms including Quantum CTEK and GuoChuang Software surged as much as 20 percent over the following two trading days, as investors bet that Beijing would answer Washington’s move with its own state backed push, a reminder that quantum computing is now traded, quite literally, as a proxy for great power competition.

Where the Science Might Actually Pay Off First

Strip away the stock tickers and the clearest near-term value sits in chemistry. Researchers at Riverlane and the pharmaceutical company Astex modeled the quantum resources needed to simulate a covalent protein-drug complex involving the cancer drug ibrutinib and found that recent advances in quantum phase estimation algorithms had cut the estimated computing time from over 1,000 years down to a few days, assuming access to an error corrected quantum computer that does not yet exist at the necessary scale. It is exactly the kind of result that captures both the promise and the catch: a genuine algorithmic breakthrough, contingent on hardware still years from being built.

Optimization is the other area already producing revenue rather than press releases. D-Wave’s annealing systems, better suited to logistics, scheduling and financial modeling than to general purpose computation, posted full year 2025 revenue growth of 179 percent, albeit from a small base, and the company’s roadmap leans into that niche rather than competing head on with the gate based machines built by IBM, Google and IonQ. The company’s newest Advantage2 processor, with more than 4,400 qubits arranged for wide connectivity, claims speedups as high as 25,000 times on certain materials science workloads, though those figures come from the company itself rather than independent peer review.

Materials science and energy storage, batteries, catalysts, superconductors, are frequently cited as future quantum applications, and Microsoft’s Azure Quantum Elements platform already layers chemistry and materials workflows on top of classical high performance computing as a kind of preview product, blending quantum-inspired methods with genuine quantum hardware access. But concrete, independently verified results remain scarcer than in pharmaceutical chemistry, and most researchers place useful, at-scale materials simulation late in the decade or beyond, well behind the narrower drug discovery use case.

China’s Bet, and the New Quantum Internet

China has treated quantum technology as a strategic priority for longer than most Western observers appreciated. Its incoming Fifteenth Five Year Plan, covering 2026 through 2030, names quantum technology alongside six other fields as a new driver of economic growth, and the country has built a state directed ecosystem linking the Chinese Academy of Sciences, state owned enterprises and elite universities, an estimated $15 billion in cumulative government and private investment, layered atop a separate $138 billion technology fund introduced in 2025.

Grimes, a technology analyst who studies China’s innovation ecosystem, told an interviewer that the country’s commitment to the field is unusually deep even by the standards of state led industrial policy.

“In China, quantum is a true number one priority.”
Grimes, technology analyst, in an interview with CKGSB Knowledge

China’s most tangible advantage may not be in raw computing power but in networking. The country has built a quantum communication backbone stretching more than 12,000 kilometers of fiber across 145 nodes in 80 cities, linked to satellites including the now retired Micius, which demonstrated the first intercontinental quantum encrypted video call between Beijing and Vienna in 2017 and later extended secure links to Russia and South Africa. Physicist Charles Clark of the Joint Quantum Institute at the University of Maryland, watching the early demonstrations, called the achievement striking in its own right.

“It’s a spectacular demonstration.”
Charles Clark, physicist, Joint Quantum Institute, University of Maryland

China aims to extend that network into a global quantum communication service by 2027, reportedly prioritizing links among fellow BRICS nations. The United States has begun building its own answer at a smaller scale: EPB, the utility formerly known as the Electric Power Board of Chattanooga, is partnering with IonQ on the nation’s first combined quantum computing and networking hub, expected online in early 2026, alongside smaller networks in New York and New Mexico. None of the American projects yet approach the scale of China’s fiber and satellite backbone.

Europe and Canada are running their own, smaller versions of the same race. The European Union’s Quantum Flagship program continues to fund research, startups and cross-border collaborations, and the bloc has moved to expand its EuroHPC mandate to build out a dedicated quantum computing pillar alongside its artificial intelligence infrastructure. Canada, for its part, has committed roughly $360 million toward quantum computing research, a modest figure next to Washington’s and Beijing’s spending but enough to keep Canadian institutions, including the University of Waterloo’s quantum research programs, competitive on the science even if not on industrial scale.

The Optimists, the Skeptics and Everyone in Between

Ask the industry’s own leaders how far away real usefulness is and the answers diverge sharply. Niccolo de Masi, IonQ’s chairman and chief executive, has told investors that demand is accelerating across artificial intelligence, defense and health care applications, with revenue expected to double again this year, and argues that Q-Day itself is arriving earlier than governments have planned for, precisely because of the pace of private investment his own company is helping drive.

“Quantum is coming sooner than people think.”
Niccolo de Masi, chairman and chief executive of IonQ

Google’s head of quantum computing, Hartmut Neven, has staked out an even bolder public timeline, telling reporters he expects real world commercial applications within five years, a direct challenge to Nvidia chief executive Jensen Huang’s widely quoted estimate that practical, general purpose quantum computing remains fifteen to thirty years away, a gap in professional judgment, between two people building the infrastructure this technology depends on, that is itself a data point about how unsettled the field remains.

Aaronson occupies the uneasy middle. He has spent his career arguing that both extremes are usually wrong, that quantum computing is neither science fiction nor a con, and that the honest position requires sitting with genuine uncertainty longer than most funding cycles or news cycles are built to tolerate. His own writing after the 2029 warning made clear that he still expects the field’s biggest open question, whether large scale, fault tolerant quantum computation is achievable at all, to be resolved empirically within his lifetime, one way or the other.

That is, in the end, the most honest answer available to the question this piece set out to ask. Quantum computers today can do things no classical machine can replicate, on narrow, carefully chosen problems that mostly do not matter yet. They cannot break a bank’s encryption, design a new drug end to end, or forecast the climate. What they can do is navigate a helicopter without GPS, distribute an unbreakable key between two cities eight thousand kilometers apart, and persuade the governments of the United States and China to treat a technology with almost no revenue as a matter of national survival.

Back at his desk in Austin, Aaronson has kept writing. Not because he suddenly believes the hype merchants were right all along, but because the pathway that once looked merely theoretical now looks, in his own careful phrase, credible. Whether that credibility arrives on schedule in 2029, on Washington’s official schedule sometime after 2033, or later still, may matter less than the fact that almost nobody serious is still betting it will not arrive at all.

Stay ahead of the stories shaping our world. Subscribe to Impact Newswire for timely, curated insights on global tech, business, and innovation all in one place.

Dive deeper into the future with the Cause Effect 4.0 Podcast, where we explore the ideas, trends, and technologies driving the global AI conversation.

Got a story to share? Pitch it to us at info@impactnews-wire.com and reach the right audience worldwide