Is Bitcoin on Borrowed Time?

My semester has been going well so far, and this year is shaping up to be a good year for publications for me with some interesting projects coming down the pipeline. It’s also been an exciting time to be following developments in quantum computing more generally.

Two big announcements this week have got a lot of attention, reaching beyond just the quantum computing community. Both Google’s announcement and [Caltech/Oratomic] have dramatically reduced the expected resource requirements to run Shor’s algorithm to break the elliptic curve cryptography that underpins Bitcoin and many other cryptocurrencies. Speculation abounds as this shifts what is known as Q-Day, the day when quantum computers become capable of breaking widely used cryptographic schemes, ever closer. To understand the timeline here it’s helpful to think about two frontiers meeting in the middle somewhere. On the one side we have the hardware, improving quality, reducing error rates, increasing qubit counts, coherence times and connectivity. On the other side we have the algorithms and software, becoming more efficient and reducing the overhead required to implement complex quantum operations. Exactly when the two will converge to make practical quantum attacks feasible is still uncertain, but these announcements suggest it may be sooner than previously anticipated. There has been some recent contrarian speculation, driven in part by works such as this which makes rather gradiose claims of quantum gravity related effects preventing large enough to be practical quantum computers. The author, Oxford’s Prof. Tim Palmer, has recent work on an alternative theory of quantum gravity that I personally find rather unconvincing, but you be the judge!

Google’s announcement was concealed behind a zero-knowledge proof to responsibly disclose the quantum vulnerabilities without exposing them to malicious actors, which is an interesting way to use cryptography! Google suggests that as few as 1,200 logical qubits, implementable with less thank 500k physics superconducting qubit, is all that is required. Oratomic’s announcement was a far more dramatic reducting to 10k-20k physical qubits needed. One potential reason here is that Oratomic’s approach leverages more efficient error correction and qubit connectivity, which significantly reduces the overhead compared to Google’s superconducting qubit implementation. From a glance it would seem that this is more evidence of neutral atoms being the superior platform for scalable quantum computing, due to their connectivity and error correction advantages.

What I’m Taking Away

Given that Google have also recently announced they are getting in the neutral atom game, I suspect that neutral atoms will continue to gain traction as a leading platform for scalable quantum computing. While Q-Day is still not here yet, it does seem prudent to make the switch to quantum-resistant cryptographic schemes sooner rather than later. Such a switch for organisations is relatively cheap and easy, no special hardware or fancy consultants needed, just following NIST’s guidelines on post-quantum cryptography with your in-house IT department should suffice, although it will still require careful planning and execution to ensure a smooth transition.