Are Quantum Computers Going to Break Bitcoin?
Back in 2017 one of the first LinkedIn articles I wrote was on Quantum Computing and the potential impacts for the finance industry (I know, I’ve been a nerd for a while 🤓 ). Whilst a LOT has happened across the quantum field in the 5 years, it’s still a useful explanation of how the technology works, so if you don’t know your qubits from your quantum entanglement then take a read
https://www.linkedin.com/pulse/power-quantum-within-financial-services-industry-tara-annison/
However one question I get asked often (twice last week) is whether the crypto industry should be worried about quantum computers.
Here’s my quick take on this ….
Essentially no, not yet, plus if it’s going to be a problem for us then it’s also going to be a problem for TradFi, national defence security, email security, and pretty much every aspect of online life which is built from public-private key cryptography!
Let’s dig into why I say this:
🔧 Specific not General
When we think about quantum computers it’s important to recognise that you have to build them for a specific purpose, more than this, they are built for a specific calculation! So it’s not accurate to imagine a Deep Thought style device which is capable of solving/breaking/computing anything and everything. This is important because it means that to be a ‘problem’ for crypto, we would need a quantum computer to be built to work on the specific algorithms that underpin crypto e.g the SHA256 algorithm which is used within the mining process, and used within the elliptic curve cryptography that generates public keys from private keys. However there’s arguably much bigger (and valuable) problems to be solving: medical calculations to map genomes, tracking environmental data to help fight climate change, and fraud detection for the trillion-dollar finance industry. Crypto right now has a market cap of ‘just’ $850billion … pennies in comparison to these opportunities! So even though funding into quantum computing research and applications doubled from 2020 to 2021 with $1.4billion, there’s still a limited pot of money and talent to research into the myriad of fields we know quantum computing could be valuable in.
⚛️ Not Enough Qubits
2021 was a monumental year for qubit count — we broke the 100 qubit boundary! To put this into perspective, when I first wrote about quantum computing back in 2017 the most advanced quantum computer from IBM had just 5 qubits!
In 2022, using the new Osprey machines, IBM hit a record breaking 433 quibits and has ambitions to reach 4,158 by 2025. We can see from their development roadmap just how much progress has been made:
However, what does this really mean in practice, should we get worried that we’ll likely break the 1,000 qubit boundary next year?
Scientists from the University of Sussex in the UK estimate that quantum computers would need at least 1.9billion qubits to crack SHA256 in a 10min window (the time it takes on average to mine a Bitcoin block) and 317million qubits to break it in an hour (the 6 block confirmations it takes to be mathematically sure a Bitcoin transaction has ‘cleared and settled’).
Extrapolating from the IBM progress that the number of qubits may triple every year (which is certainly a tall order and I’m sure even the more optimistic quantum scientist would have a cold sweat at that target!) it would still take us until 2136 until we get over the 317m qubit threshold and 2038 until we hit the 1.9billion qubit barrier.
There’s also many challenges when we start increasing the qubit count — one notable one which is shared with proof of work mining in crypto is heat — and generating too much of it! As you increase the number of qubits you have to manage the heat that’s generated from the microwave pulses used to control them. This means increasing costs for cooling technology and likely many of the same political/moral/environmental challenged when it comes to the tradeoff in energy use vs benefits.
Another important factor when increasing qubits is cost — it’s estimated that a single qubit costs around $10k and so when we’re thinking of systems in the millions and billions of qubits there’s certainly a question mark around the cost to build and maintain these systems. Especially when the ‘prize’ for doing so is hacking a network (e.g Bitcoin) which has total market cap of ‘just’ $326billion and which will likely see this plummet to zero if a quantum-exploit was made.
It’s also important to note that just raw qubit numbers aren’t a sufficient measure of a quantum computer’s power. Instead we need to think about: speed of calculations, the number of qubits and the quality of the data. Enhancing all 3 are important if we wish to harness the power of quantum and there’s whole areas of quantum research which look at either one of these elements, plus sub-divisions thereof.
🧮 Quantum Proof Research
However that also does not mean that crypto will be broken at the point we have a sufficient number of qubits. This is because there’s already ongoing work into quantum proof algorithms — basically maths that can’t be broken even with quantum computers!
The US National Institute of Standards and Technology (known as NIST) have been researching quantum proof algorithms since 2016 (and likely investigating them far before this) and in July of this year, after numerous rounds of testing and peer review, announced four quantum proof algorithms which will now be further battle tested and implemented.
The algorithms are:
- CRYSTALS-Kyber for general encryption due to its speed and small encryption keys
- CRYSTALS-Dilithium, FALCON, and SPHINCS+ for digital signatures
As such, whilst there’s no immediate need to be swapping out SHA256 for Bitcoin, if the time came then since Bitcoin is software and requires a consensus change for an update, as long as the network participants agreed (and it would of course be in their interest to) then Bitcoin could switch to a new quantum-safe algorithm instead. Therefore protecting it against the dangers of advancing quantum systems. What’s perhaps interesting to note here is that the nimbleness and open source nature of Bitcoin potentially means it will be able to adapt to this risk much faster than TradFis and complex financial institutions which are built from multi-layered and incredibly complicated systems, and for which deploying a change such as this would likely require rounds of stakeholder engagement, risk approvals, board meetings and even more research into what part of the system does what!
Originally published at https://www.linkedin.com.
