Talk to someone for any length of time about Bitcoin, and the question of scaling arises. From a technical point of view, scalability is not a challenge that is unique to cryptocurrencies, or to blockchain-based applications. Any application or network that is designed for high traffic or multiple concurrent users needs to be programmed and configured in such a way that many people can access it simultaneously without degradation of performance.
Public proof-of-work blockchains pose specific challenges because energy-intensive transactions must be recorded and consensus agreed on thousands of different computers to ensure the integrity of the network’s history. Detractors often cite Bitcoin’s throughput limits – it can process about seven transactions per second compared with around 65,000 per second on the Visa network – as a total failure of its original purpose.
Throughout Bitcoin’s twelve-year history, the debate around how – and whether – to increase its capacity has become a major topic, even prompting a standalone conference, Scaling Bitcoin. However, equating Bitcoin to something like the Visa network is not a like-for-like comparison. As recent institutional acquisitions of Bitcoin indicate that more and more people are focusing on the original cryptocurrency as a store of value rather than a high-volume, consumer-focused payment network, it is probably true to say that it is Bitcoin’s function as a global settlement system that is valuable, as well as its ability to allow people to secure their own wealth in an uncensorable way.
In simple terms, we should not judge Bitcoin for its ability to process $2 dollar payments for cups of coffee quickly and cheaply, but instead think about the ways that its blockchain can support such payments in an indirect way. The best way to think about this is that we should look at the scalability of the entire architecture rather than the scalability of the baseline protocol itself.
Hence, alongside purely technical solutions such as the ones I describe below, we should also think about the idea of institutional scaling, as Nic Carter suggests here: Solutions such as Liquid, a Bitcoin-backed sidechain that allows rapid transactions for traders and exchanges as well as a host of other features, fall into this category.
Over the last decade, approaches to scalability have focused on different solutions: here, we look at a brief history of approaches that have been tried, and what is being planned for the future.
When we talk about changes to Bitcoin’s software, it is important to remember that the whole network runs on community consensus. While changes to the core software are carefully reviewed for quality and changes extensively debated, anyone can create a fork of the code, release it into the wild and allow miners to run it.
By 2014, some developers were advocating that the size of the blocks that contain Bitcoin transactions should be increased from 1MB, to allow for a greater throughput of transactions. Eventually a fork of the software was released which allowed for this (called Bitcoin XT) but the majority of miners opted to continue using the original software.
Continuing pressure on the network meant that scalability problems remained high on the agenda, and by 2017 a hard fork was implemented which became known as Bitcoin Cash. It had the backing of some prominent Bitcoiners, including Roger Ver, and the move prompted bitter disagreements on social media, with both sides claiming the right to the Bitcoin name. The maximum size of a Bitcoin Cash block was 8MB (later increased to 32MB), and a further fork, known as Bitcoin SV, implemented a block size cap of 128MB.
Meanwhile, improvements to Bitcoin’s core code meant that capacity (the number of transactions that could be fitted into a block) could be increased. An improvement called Segregated Witness was introduced in 2017 after first being presented at the end of 2015. SegWit, as it was abbreviated, is a protocol upgrade that reduces transaction size rather than increasing block size by removing signature data from the block of data that needs to be processed in each transaction and having it attached separately.
The changes described above all involve making changes to Bitcoin’s software. Other scaling solutions involve creating separate protocols that are built on top of the main network and which can do a big chunk of the work that would otherwise happen on the main chain.
This means that these so-called second layers can offer high throughput and can be suitable for much smaller transactions. In the very early days of Bitcoin, for example, people talked about using the network for microtransactions of a few cents because congestion and transaction fees were so low. Bitcoin’s best-known Layer 2 network, Lightning, is now used for these very small transactions.
While Lightning is very much a product under development, its potential for scaling payments on the Bitcoin network in future is undeniable. When two nodes on the network create a channel by each sending some Bitcoin to a specific address, this creates a channel where payments can go backwards and forwards between them without the transactions being broadcast to the main Bitcoin blockchain. Eventually, if one party decides to terminate the channel and reclaim their Bitcoin, the transaction is broadcast and written on the main chain. Payments that have happened in the interim are known as off-chain payments.
As specified on the lightning.network website: “Transactions can be made off-chain with confidence of on-blockchain enforceability. This is similar to how one makes many legal contracts with others, but one does not go to court every time a contract is made.”
Blockstream’s Liquid is mentioned in the intro as a network for allowing exchanges and market-makers to make settlements quickly off-chain. Like other sidechains, Liquid provides a scaling solution by allowing users to move funds between itself and the main Bitcoin network, while enabling transactions on its own blockchain (blocks are generated here every 10 minutes). Liquid itself is secured by a consensus mechanism called Strong Federation, which obviates the need for a costly Proof of Work consensus.
Liquid is far from the only sidechain: one of the most interesting is Rootstock (RSK) – ‘DeFi for Bitcoin’ – which was initially proposed in 2014, and allows Turing-complete smart contracts to be executed secured by the Bitcoin network. Rootstock claims Bitcoin-style security with Ethereum-style flexibility, and enables rewards for miners via a process called merged mining (using their computational power to mine blocks on BTC and RSK simultaneously so that the work done on one blockchain can be used as valid work on the other).
CommerceBlock is also creating an interesting ecosystem around Bitcoin, including a statechain implementation known as Mercury. Defined in their own words:
“a State Chain is a Bitcoin second-layer protocol in which instead of spending an old UTXO and creating a new UTXO as in a base layer transaction, ownership of coins can be transferred by directly handing over the information necessary to spend a UTXO through some other communication channel, while ensuring that the sender becomes unable to spend the UTXO themselves once the transfer is complete.”
Because statechains allow users to make transactions off-chain, outside of blocks, they don’t have to wait. In this instance, statechains are powered by multi-signature transactions, which require that more than one user sign off on a transaction before it can be completed.
Smart contracts – sets of rules written in software that specifies conditions around a particular transaction or transactions – are a hugely important feature of blockchain technology, as they allow the automation of various processes and agreements without having to trust intermediaries.
When we think about networks that support smart contracts, Ethereum is probably the first example that springs to mind, but some of the first blockchain-based smart contracts were written on protocols such as Mastercoin (later rebranded as Omni) and Counterparty – who described themselves rather poetically as ‘writing in the margins of Bitcoin transactions’ – which were built on top of the main Bitcoin blockchain.
Protocols that extend the functionality of Bitcoin with smart contracts are indeed a form of scaling as they allow for more complex transactions that can do more things. One of the most exciting and long-awaited changes to Bitcoin’s core code that will be activated in 2021 is Taproot, which will allow for more lightweight smart contracts that are also more private. Taproot uses Schnorr signatures, which are smaller and faster. This piece in Bitcoin Magazine provides an excellent overview:
“The removal of middlemen leads to lower fees, faster deal execution, free market exposure, larger potential investor base, automated service functions, and lack of financial institution manipulation.”
Can Bitcoin scale? The answer is yes, but the question is very much, how? The next year or so will be interesting.
As we described in the introduction, scaling is a challenge for any blockchain that relies on proof of work. Congestion and high transaction fees on the Ethereum network, on which many DeFi protocols and NFT trading and minting depends means that there is much work being done on scaling Ethereum – and this will be the focus of the next article in this two-part series.