Latest / Quickly Quantum / IonQ's Bitcoin-Cracking Math: Real Physics, Marketing Slide — September 10, 2026
Transcript
- Twenty-five point seven days — that's how long a future quantum computer would need to crack the encryption underneath every Bitcoin wallet on Earth, according to a new paper out of IonQ. Is that the countdown clock crypto's been dreading, or a really good marketing slide? That's our main story today. Before that, in the headlines: IQM lays out a path to logical qubits for Finland's national AI supercomputer, IBM and Lockheed Martin open a quantum hub in Switzerland, a chemistry benchmark gets dramatically more accurate, and the US Treasury starts prepping banks for the quantum threat. Welcome back to Quickly Quantum, your daily brief on the quantum frontier. It's Thursday, September 10, 2026. Let's get into it. Finland's making a long bet on error correction. IQM announced a staged upgrade path to deliver LUMI-IQ — a quantum computer built from day one for logical qubits, the more stable qubits you get by combining many noisy physical ones through error correction — to Finland's EuroHPC-funded LUMI AI Factory. This isn't one machine landing tomorrow, it's a three-step rollout. Step one, in 2027: an IQM Halocene H4 system with a hundred fifty physical qubits and early error-correction. Step two, in 2028: an upgrade that lowers the logical error rate and adds real-time error correction. Step three, in 2029: the Halocene H5, which IQM says will support up to nine logical qubits, with lattice surgery and T-gate teleportation — the operations you need to actually run programs on error-corrected qubits, not just hold them steady. The whole thing is jointly funded by the EuroHPC Joint Undertaking along with Finland, Czechia, Norway, and Poland, and it'll be hosted at CSC's data center in Kajaani. IQM's CEO, Jan Goetz, said the setup lets Europe 'shape its own direction' in quantum rather than just renting access to someone else's machine. Here's the catch: nine logical qubits in 2029 is three years and two hardware generations away from the hundred fifty physical qubits landing next year — those are stages of the same rollout, not two options you pick between today. Real-time error correction at that scale hasn't been demonstrated by anyone yet. Still, a national government-backed customer signing onto a phased fault-tolerance roadmap is a real vote of confidence in where the field is headed. Switzerland's getting its first IBM quantum computer, and it's arriving through a defense deal. IBM, Lockheed Martin, and Swiss defense procurement agency armasuisse announced a quantum innovation hub at ETH Zurich, anchored by an IBM Quantum System Two running on the company's newest processor, Nighthawk. It's expected to be operating at the Swiss National Supercomputing Center in Lugano by the end of this year. The research agenda leans practical — quantum sensing for navigation and improving additive manufacturing of metal alloys, both areas Lockheed already works in. Swiss startups, academia, and industry get access too. Lockheed's chief technology officer, Craig Martell, framed it as extending the company's long relationship with IBM into quantum and AI. That structure matters, though: this is a defense-offset agreement tied to Lockheed's Swiss procurement relationship, so Switzerland is essentially getting a quantum computer partly because Lockheed owes the country something on a separate defense contract. On the chemistry side, Qedma says it just made quantum computers more trustworthy for real chemistry problems. Error mitigation is the workaround for noisy hardware: instead of building perfect qubits or running the heavy error-correction codes IQM is chasing, you clean up the noise in software after the fact, so you don't have to wait for fully fault-tolerant machines to get useful answers. University of Copenhagen's Stephan Sauer said the study let his team explore how far error mitigation can push accuracy on today's hardware. It's a number that came from Qedma itself, by Quantum Zeitgeist's account — one more data point in the race among error-mitigation vendors to show real gains on real chemistry problems. Meanwhile in Washington, banks just got a homework assignment. Quantum Zeitgeist reports that the Treasury Department has stood up a Quantum Readiness Task Force, following an executive order, to coordinate the financial sector's shift to post-quantum cryptography — cryptography designed to resist a future quantum computer, unlike today's encryption, which quantum computers could eventually break. According to that reporting, the task force splits into three workstreams: sector alignment on the transition, third-party and vendor readiness, and digital-assets and emerging-technology risk, which means banks but also crypto exchanges and tokenized assets are on the list. Treasury's Luke Pettit was quoted saying the goal is a transition that's 'coordinated, risk-based, and operationally resilient.' Quantum Zeitgeist's account doesn't spell out hard deadlines or enforcement mechanisms for banks that drag their feet. But the timing is pointed: this lands the same week IonQ put an actual number on how a quantum computer could crack Bitcoin's encryption, exactly the scenario this task force exists to get ahead of. Call our main story today the Twenty-Six Day Question: is this the moment the quantum-crypto countdown became a real deadline, or is it still a paper exercise dressed up as news? IonQ published what it's calling the first complete, end-to-end fault-tolerant resource estimate for running Shor's algorithm — the quantum algorithm that, in theory, can factor the giant numbers underlying modern encryption — specifically against secp256k1, the elliptic curve that secures every Bitcoin signature. According to The Quantum Insider's reporting on the paper, the headline numbers are nineteen thousand three hundred ninety-seven physical qubits, and twenty-five point seven days of runtime to break it. Now, here's what makes this different from the usual quantum-threat headline. Most of these estimates hand-wave the hardest parts — the actual error-correction overhead, the real gate counts — and approximate their way to a scary number. IonQ says it compiled this one all the way down to the physical error-correction level, meaning the qubit count and the runtime aren't back-of-envelope guesses, they're the output of an actual circuit compilation using the company's Walking Cat architecture, its design for networking trapped-ion qubits together. So does a working machine exist that could actually do this? No. IonQ's own roadmap — the one it sells to customers and investors — targets a fully fault-tolerant machine with ten thousand physical qubits only in 2027, with further advances promised in 2028. That's roughly half the qubit count this attack needs, arriving next year at the earliest, on a roadmap written by the same company that published the paper claiming the attack is possible. What this paper actually does is turn a vague fear into a specific, checkable number — the runway to an actual attack machine hasn't gotten any shorter. The timing helped it travel: CoinDesk and blockhead.co both ran pieces this week framing Bitcoin's quantum exposure as newly urgent, and it landed the same week Washington finalized three hundred million dollars in CHIPS Act funding for other quantum hardware makers, a story we covered on the show yesterday. One important distinction: the post-quantum signature schemes already being rolled out, ML-DSA and SLH-DSA, are explicitly unaffected by this class of attack — this is about breaking today's Bitcoin signatures, not tomorrow's replacements. IonQ made its own case loudest. On X, the company's official account posted, quote: 'The first complete, end-to-end fault-tolerant resource estimate for running Shor's algorithm — the team used IonQ's Walking Cat architecture to demonstrate how a specific application can be optimized for a trapped-ion quantum computer. This work concludes that a 20,000-physical-qubit IonQ quantum computer, once developed, is expected to be able to break secp256k1, the 256-bit elliptic curve used by blockchain technology such as Bitcoin, in just under 26 days,' end quote. And that's the tension worth sitting with: IonQ is simultaneously the source of this result, the peer-reviewable science behind it, and the company whose stock moves every time 'quantum' and 'Bitcoin' share a headline. None of that makes the compilation wrong — end-to-end resource estimates are real, publishable science, and this one appears to go further than prior estimates by not skipping the hard error-correction math. IonQ gets to publish a genuine technical achievement and a recruitment pitch for its own roadmap in the same document, and the headline writers downstream mostly ran with the second part. We've said before on this show that the multi-trillion-dollar quantum-crypto risk numbers floating around are exposure estimates, not countdown clocks — nobody credible claims a cryptographically-relevant quantum computer exists today. This paper sharpens that read rather than overturning it: the gap between precisely simulating what an attack would cost and actually building the machine that runs it hasn't closed. IonQ's own roadmap puts its fully fault-tolerant machine at ten thousand qubits in 2027, half of what this attack needs, with further advances still required after that. There's also a narrower technical question the paper doesn't answer: does IonQ's Walking Cat architecture generalize to other cryptanalysis targets, or was secp256k1 — Bitcoin's curve, guaranteed to generate headlines — chosen because it travels well in the press? So where does this land? My read: this is careful engineering work that happens to arrive wrapped in the most clickable possible framing. The number — twenty-five point seven days, nineteen thousand three hundred ninety-seven qubits — is a genuine technical achievement in compilation, the first time anyone's walked Shor's algorithm all the way down to physical error correction for a specific target. The theory just got sharper; nothing about your Bitcoin holdings is less safe today than it was last week. If you build post-quantum migration policy for a living, like the Treasury task force we just covered, this is exactly the kind of concrete number that should feed planning, not panic. Time for the Hype Check. I'm putting this at a 6. The compilation work is genuinely more rigorous than the usual quantum-threat estimate, and that's worth something real. But the headline framing — twenty thousand qubits, twenty-six days — is doing work the underlying roadmap can't back up for years, and it's coming from the one company with the clearest financial incentive to make Bitcoin sound like it's on a clock. So the next time you see a headline with a specific qubit count and a doomsday date attached, the question isn't whether the math checks out — it's how many years stand between the paper and the machine that could actually run it. If today's episode helped untangle that, follow Quickly Quantum wherever you listen, so it's ready for you tomorrow. This has been Quickly Quantum, an AI-voiced podcast, created and built by a real human using today's cutting-edge technology. Nothing you heard on this show is financial advice. I'm Brian Lampert, and I'll catch you all tomorrow — take care! I also host Space Stakes: the business of the new space race, every day. What actually flew, what the contract is really worth, and who has customers. Find it wherever you get your podcasts.