Is quantum computing going to break the blockchain?

At a cryptocurrency seminar last year, one of my favourite moments was when someone asked what happens when quantum computers become reality? Dead silence and a long pause before the speaker said something to the tune of “we will figure that out when we get there”.

I think I need to explain that a little bit. The entire premise of blockchain technology and the cryptocurrencies, NFTs and smart contract built on that is that the decentralised ledgers are immensely secure and cannot be hacked with modern day computers. I am oversimplifying a bit, but in order for a transaction to be accepted by the blockchain, 51% of the computers in a network that share the blockchain need to agree that the computer that claims to be the new rightful owner of a cryptocurrency or any other asset need to agree that this computer is indeed the legitimate owner. And the network only accepts claims that have a proof of work (PoW) which is essentially a massive multiplication exercise of several very large numbers (as I said, I am oversimplifying here, so don’t complain). Once such a PoW has been submitted to a blockchain network and more than 50% of computers accept it as correct, a new block will be added to the chain and the longer blockchain will be considered the true blockchain. Submitting a PoW for a new cryptocurrency is what brings a new token/coin into existence. Similarly, by submitting PoW, contracts can be brought into existence that prove ownership of certain assets without having to rely on central databases or potentially corrupt government officials.

Now imagine you can do this PoW faster than all the computers in a network put together. If you could do this, you could constantly outrun the verification process and generate new blocks in the blockchain faster than the rest of the network can check them. This means that you can add new blocks to the blockchain without these blocks being verified. And since the all blockchain technology is set up to assume the longest blockchain is the true one, you can essentially “hack” the system by creating a longer blockchain than any other in the network which means that all the other computers simply accept your blockchain as the correct one against which to compare any new proof of work.

With modern computing power it is simply impossible to create such a 51% hack. But quantum computers will be so much faster that at some point they will easily be able to outpace any network of traditional computers. In fact, one of the most important things to understand about quantum computers is that they are not just faster computers. Traditional computers are based on transistors that allow to differentiate between two states (called “bits”), 0 and 1.

Quantum computers, meanwhile can take on both 0 and 1 at the same time and superimpose these “Qbits”. If that sounds weird, think of a typical old-fashioned computer that encodes letters or numbers as a series of 8 bits. there are 256 different characters or numbers that can be coded with these 8 bits and at any given time a transistor in a traditional computer will be in one of the 256 possible states. But a quantum computer with 8 Qbits will be able to take all 256 states at the same time and use them for computations simultaneously. You see how the advantage of quantum computers grows exponentially as they include more Qbits?

This means that algorithms in quantum computers have to be completely redesigned from scratch in order to take advantage of these computational abilities. But it also means quantum computers are so much more powerful that they can easily crack any problem that with traditional computers would be unsolvable within the remaining lifetime of the universe.

So, assume you are the first person or company in the world to build a fully functioning quantum computer. Since all the networks in the world are based on traditional computers, you could take over all the blockchains of the world within seconds. Only once the majority of the computers in a network become quantum computers as well is the blockchain safe again. But by then, it may be too late.

This benefit of quantum computers has been shown to hold even when these computers haven’t really achieved what is called a true quantum advantage, i.e. the quantum computer can solve problems that are impossible to solve for any traditional computer. All we need is that it takes traditional computers longer than any reasonable time period to solve the same problem a quantum computer can solve in seconds and all the blockchains in the world have become unsafe and can easily be hacked by whoever has access to the first quantum computers. In other words, when quantum computers become reality, blockchain technology has to be completely redesigned from scratch or it loses all its advantages in terms of decentralisation and security.

But quantum computers are science fiction, aren’t they? Yes, they are, but they are being developed right now. And if you extrapolate current advances in the speed of computing according to Moore’s Law into the future, a single quantum computer could hack the Bitcoin blockchain by 2045.

Lower estimate when quantum computers can overtake the hash rate of Bitcoin

Source: Bard et al. (2021)

But that is a low estimate because it assumes two things. First, it assumes that progress in quantum computing advances at the same rate as progress on traditional computers. But we know that a new technology progresses much faster than a well-established technology. Second, the chart below uses the Bitcoin blockchain which is by far the most complex and computationally intensive one (which is why Bitcoin does is not able to compete as a payment system with existing payment systems like PayPal or credit cards). Other blockchains like Ether or blockchains underlying commercial applications use much smaller blockchain networks. And according to a new study on quantum computing advantages this could happen as early as 2023.  

Personally, I think 2023 is not a realistic date, but as someone who is reading more and more about advances in quantum computing, I wouldn’t be surprised if we reach this critical junction within this decade. And that would still mean that all blockchain applications would have to be fundamentally redesigned in the next couple of years or risk becoming unsafe and essentially useless.