Quantum computing, long grouped with controlled nuclear fusion and artificial general intelligence as technologies that always seemed to be a decade away, is now being presented as the one that may arrive first.
In an article published by MarsBit and credited to the WeChat account Zimu AI, author Miao Zheng wrote that commercial-grade quantum computers have already entered cloud data centers, while the practical threshold of 100 logical error-corrected qubits could be crossed as soon as next year.
If that happens, the article says, quantum systems could reshape drug discovery, materials science and financial modeling. For crypto, the tone is very different. The report argues that elliptic curve cryptography, which protects private keys for Bitcoin and Ethereum, sits directly in the path of the kind of attacks quantum machines are expected to handle well.
Why 100 logical qubits matter
The article cites writer Thomas Black, who said quantum computing remains in a “pre-ENIAC” stage, but that a major milestone is close: logical qubits.
It uses ENIAC, the programmable general-purpose electronic computer introduced in 1946, as a reference point. ENIAC weighed about 30 tons, used 18,000 vacuum tubes, filled an entire room, performed only 5,000 additions per second and would cost about $9 million in today’s terms, according to the article. A standard laptop now handles billions of calculations per second.
Classical bits are either 0 or 1. Qubits can exist in superposition and can become entangled with one another, letting them carry more information. The problem is fragility. Vibrations, heat and even a passing cosmic ray can collapse a quantum state and introduce errors, the article says.
That is why the industry relies on redundancy. Information is spread across many physical qubits, and error correction is used to synthesize logical qubits. The article compares this to 100 workers, each remembering only a fragment of a larger record, with cross-checks in place to identify and correct any one person’s mistake.
Black said quantum computing needs 100 logical qubits to become commercially useful. Crossing that level would let machines tackle problems that classical systems struggle with, including molecular simulation, new materials design and optimization of complex systems. In his view, the best-case scenario is reaching that mark next year.
Progress from Quantinuum, IBM and Infleqtion
The article says Quantinuum is among the leaders. Its Helios machine has 98 physical qubits, has encoded 48 logical qubits and delivers 99.921% two-qubit gate fidelity.
On Aug. 11, Quantinuum announced a partnership with Oracle to place Helios in a U.S. Oracle Cloud Infrastructure AI data center. The article describes this as the first time a quantum system was physically installed in the AI facility of a mainstream cloud provider. Oracle expects to use the machine next year for AI training and inference.
Quantinuum’s next machine, Sol, is scheduled for release in 2027 and is aimed squarely at the 100-logical-qubit threshold.
IBM is another major player mentioned in the report. It says IBM and the University of Chicago reached 70 logical qubits in an experiment this July. IBM said the result produced by that 70-logical-qubit system could not be reproduced by the strongest supercomputers on the market, and that the quantum run took about 15 minutes.
Infleqtion, which follows a neutral-atom approach, plans to exceed 50 logical qubits next year at a quantum campus in Illinois. The company’s CTO said, “At 100 logical qubits, you begin to solve important problems in materials science and chemistry that classical computers cannot solve.”
The article outlines the competing hardware paths. IBM, Google and Rigetti use chip-based mechanical qubits, which are easier to scale but come with higher error rates. Other teams use natural particles, trapping atoms with magnetic fields and controlling them with lasers, or manipulating photons with beam splitters and phase shifters. Infleqtion uses neutral atoms, Quantinuum uses charged ions and PsiQuantum uses photons.
No one knows which route will win, the article says, and several may coexist.
It also notes that in May, the U.S. Department of Commerce awarded $2 billion in funding across nine quantum computing companies, including IBM, Rigetti, Quantinuum, Infleqtion and PsiQuantum, covering all of the major approaches.
Examples of what quantum machines can do
The report argues that quantum computing’s edge over conventional systems is starting to show up in real demonstrations.
It cites a 2025 D-Wave demonstration involving a magnetic material. The article explains magnetism through electron spins, described as tiny needles pointing up or down and constantly influencing one another while changing under temperature shifts and external magnetic fields. D-Wave’s quantum computer, it says, modeled how thousands of these spins flipped over time under quantum rules, how they interacted and what magnetic behavior emerged.
According to the article, a classical supercomputer would need millions of years for the same task, while D-Wave’s machine completed it in minutes. The paper appeared in Science and passed independent verification.
Thomas Black also said logistics companies have started using quantum algorithms in pilot projects for route planning, fleet scheduling and inventory management, with efficiency gains of 15% to 30%.
In finance, the article says JPMorgan, Goldman Sachs and HSBC have for years explored quantum computing for risk modeling, options pricing and portfolio optimization. IBM’s modular Kookaburra system has connected more than 4,000 physical qubits, while its open-source Qiskit framework now counts more than 600,000 developers.
On Aug. 24, Japan launched Shunkai, described in the article as the country’s first full-stack neutral-atom quantum computer, formally joining the race.
With technology, business deployment and capital moving at the same time, the article argues that quantum computing is shifting from “10 years away” to something that could arrive next year.
The article’s case against crypto security
The report then turns directly to cryptocurrencies. It says private keys in systems such as Bitcoin and Ethereum are, in essence, random numbers. Public keys are generated from private keys through elliptic curve cryptography, and wallet addresses are produced by hashing those public keys.
It compares the public key to a bank and the private key to a signature and password. On classical machines, deriving a private key from a public key is effectively infeasible. The article says that even the world’s best supercomputers would need more time than the age of the universe.
For future quantum machines, the report says, that process may take only hours. A classical computer attacks a private key by trying one key after another. A quantum computer, in the article’s analogy, acts more like a master key that can explore many possibilities at once.
In 1994, mathematician Peter Shor proposed a quantum algorithm that can factor large integers and solve elliptic curve discrete logarithms in polynomial time, making it applicable to RSA and ECC. The article says the common industry response for years was that the theory looked elegant but the hardware remained too distant. Once systems above 100 logical qubits exist, it argues, that changes.
The piece says Google published a paper in March arguing that breaking Bitcoin’s elliptic curve cryptography might require fewer than 500,000 physical qubits, far below the industry’s earlier estimate of about 9 million. That would cut a possible attack window from hours to tens of minutes, according to the article.
Around the same time, Caltech and startup Oratomic published work saying that, in theory, just over 10,000 physical qubits could break ECC-256 in roughly a dozen days. The threshold, the article says, keeps falling.
Because one logical qubit corresponds to dozens or even hundreds of physical qubits, the article argues that if Oratomic’s framework proves right, pushing logical qubits past 1,000 would put any cryptocurrency at risk.
Harvest now, decrypt later and transaction front-running
The report says attackers are already incentivized to adopt a harvest-now-decrypt-later strategy, or HNDL.
That means collecting large on-chain addresses, transaction histories and encrypted communications today, then waiting for mature quantum computers to decrypt them later. Since blockchain data is public and permanently stored, the article says, the cost of collecting it is close to zero.
Google’s paper, according to the report, also said about 6.9 million BTC, equal to 30% of total supply, sits in wallets where the public key has already been exposed.
The article also describes a “transaction front-running” scenario. If a quantum computer can derive a private key in minutes, while a Bitcoin block takes about 10 minutes to confirm, an attacker could broadcast a competing transfer after seeing a transaction and move the coins first before confirmation.
Under that framework, the report says blockchain’s much-cited immutability would fail against quantum attacks. Even moving funds into a so-called quantum-safe wallet would not erase old addresses and old transactions already written on-chain.
The article goes one step further and says the market may not wait for a real-world private key break. If investors come to believe that successful attacks are likely within one or two years, institutional money could leave before the first theft occurs, pushing prices lower in advance.
Post-quantum migration is already under way
The article says the quantum threat extends far beyond crypto and reaches the broader internet security stack.
It states that on June 22 this year, the U.S. signed Executive Order 14412, requiring federal agencies to migrate all high-value systems to post-quantum cryptography by the end of 2030, complete digital signature migration by the end of 2031 and bring contractors into compliance by the end of 2030.
The U.S. National Institute of Standards and Technology, or NIST, formally released three post-quantum cryptography standards in 2024, the article says. The National Security Agency has also required that newly procured national security systems support post-quantum algorithms starting Jan. 1, 2027.
Current encryption methods such as RSA and ECC depend on hard mathematical problems, including integer factorization and elliptic curve discrete logarithms. Post-quantum systems replace them with problems that quantum computers are also expected to struggle with, including lattice-based cryptography.
Google has set 2029 as its own internal deadline for completing migration. The gap, the article says, is that defensive migration may take five to 10 years while the offensive breakthrough could arrive in one to two years.
That delay exists because this is not just a password swap. It involves certificate systems, software signing, supply-chain trust and wholesale replacement across IoT deployments. Any place that still relies on older cryptography remains exposed.
Once the quantum era arrives, the report says, neglected devices could become permanently exposed nodes on the network. Attackers would not need to break a main server if they can find a device in a corner that has gone unpatched for a decade and use it as a jump point.
The article says large enterprises may need 12 to 24 months simply to inventory where and how cryptography is used internally. Embedded devices that cannot receive firmware updates are an even larger problem, since they may need to be retired or kept in service with known risk.
HNDL makes the timeline more urgent. Data moving through encrypted channels today may already be stored by adversaries for future decryption, the article warns.
G7 guidance, Europol and SandboxAQ
Even an organization with Google’s resources and incentive structure still targets 2029 for completion, the article notes, which suggests an even steeper burden for ordinary companies and individuals.
Earlier this year, a G7 cybersecurity expert group issued a coordinated roadmap to banks, insurers, exchanges and regulators, urging the financial sector to accelerate post-quantum preparedness through six stages.
Europol also released a report in January that offered financial institutions a framework for handling quantum threats.
The article closes by pointing to the commercial response. SandboxAQ, a company spun out of Google, focuses in part on helping enterprises and governments migrate to post-quantum systems, and its valuation has topped $5 billion. In the report’s telling, money is already being made on quantum defense before large-scale quantum computing fully arrives.