Quantinuum said this week it has reached a record 99.9995% logical fidelity in quantum error correction, a result the company frames as a step toward a commercially viable, fault-tolerant quantum computer by 2029. The announcement lands as cryptography researchers keep asking the same uncomfortable question: how much longer do the elliptic-curve signatures protecting bitcoin wallets hold up?
Error correction is the part of quantum computing that's hardest to make boring. Qubits are fragile, and the errors they throw off have to be caught and fixed faster than they accumulate. Quantinuum's number is about how reliably its error-corrected — "logical" — qubits behave, not the raw physical qubits underneath them.
Why the fidelity number matters
Raw qubit counts make for good press releases. They don't tell you much about whether a machine can run a long calculation without collapsing into noise. Logical fidelity is the more honest gauge: it measures the corrected qubit, the thing you'd actually build an algorithm on.
Pushing it to 99.9995% suggests the error-correction overhead is shrinking — each logical qubit needs fewer physical ones to stay stable. That's the direction you want if you're trying to scale. It's also the direction that makes cryptographers nervous.
The bitcoin question
Bitcoin's security rests on public-key cryptography — ECDSA for signatures, SHA-256 for hashing. A sufficiently powerful fault-tolerant quantum machine running Shor's algorithm could, in theory, derive a private key from a public one. That's not a today problem. It's a 2029-and-beyond problem, which is exactly what Quantinuum just put on the calendar.
The company hasn't claimed it can break bitcoin. It has claimed it's building the kind of machine that could eventually try.
What's actually been demonstrated
Quantinuum's release is specific about the fidelity figure and the 2029 target. It is not specific about qubit counts, runtime, or which algorithm was run to produce the number. Those details matter for anyone trying to judge how close the machine is to doing useful cryptographic work, and they aren't in the announcement.
The company's previous milestones have been peer-reviewed and replicated by outside groups, which is more than most quantum vendors can say. That track record is part of why this week's number is being taken seriously rather than filed under marketing.
The migration clock
Post-quantum cryptography exists. NIST finalized its first standards in 2024, and wallets, exchanges, and custody providers have been slow to adopt them — mostly because the threat has felt abstract. Quantinuum's 2029 target doesn't change that overnight. It does give the abstract a date.
The next concrete checkpoint is Quantinuum's own: it says it will publish the technical details behind the 99.9995% figure and submit the work for peer review. Until that lands, the number is a claim from a company with a good track record, not a verified result. The cryptography community will want the latter before it starts rewriting bitcoin's signature scheme.
