Blockchain Scaling 101
In today’s cryptocurrency market, there is a high demand for networks with great throughput. 15 transactions per second (TPS) that Ethereum processes today are not enough, to say the least — there are too many users in the network, the blockchain often gets congested, and fees go to the moon.
Newer platforms explicitly focus on scalability: Avalanche processes 4,500 TPS, Terra does 10,000, and Solana goes as high as 50,000. Older networks attempt to scale, too: Ethereum is currently on a 4-year-long journey that is supposed to raise the blockchain’s throughput all the way up to 100,000 TPS and beyond.
One of the key concerns about scaling is that higher network efficiency often goes hand in hand with reduced security and decentralization. The difficulty of maintaining all three parameters at a good level is what Vitalik Buterin once called a scalability trilemma.
How does the crypto industry address this challenge, and what approaches to scaling are there today? Let’s find it out.
Two Types of Blockchain Scaling
Scaling is about allowing more users to make more transactions, preferably without harming the network’s security and decentralization. It feels like the most apparent way to achieve this is building a more efficient blockchain or implementing changes to an existing one — this kind of approach represents on-chain solutions. Another means to increase throughput is organizing transactions in a way that allows them to optimize the traditional confirmation process. This is an off-chain approach to scaling. Let’s take a look at both.
On-Chain Solutions for Scaling
This type is often called layer one or base layer scaling, meaning that a cryptocurrency implements a more efficient protocol compared to its predecessor or makes some scaling-focused changes to its blockchain after launch. These improvements aim to let the network put more data in blocks in less time. Bitcoin Cash (BCH), Litecoin (LTC), Ripple XRP, and EOS are the networks that went this way. Bitcoin has implemented an L1 scaling solution called SegWit.
Here are the main ways to scale the network on-chain:
- Increase the amount of data that can be put into a block. Thus, a block can process more transactions. Bitcoin Cash (BCH) and Bitcoin SV (BSV) are Bitcoin hard forks that were launched with this type of protocol change.
- Accelerate block formation. Litecoin is a Bitcoin successor with 4 times faster block time (2.5 minutes against 10). In Ethereum and many other coins, new blocks are created every few seconds.
- Reducing the space transactions occupy in a block. By implementing the SegWit update in 2017, Bitcoin has decreased the amount of transaction data that had to be put into a block, so the latter could fit more transactions.
- Changing the blockchain design completely. To handle the growing number of transactions, Ethereum is currently undergoing a series of updates changing its consensus algorithm from PoS to PoW and split the network into 64 shard chains that will process transactions in parallel.
Although these are some well-tested methods to improve scalability, they have their drawbacks as well.
By increasing the block size, we also raise the power requirements for mining. You need stronger hardware and more electricity to process larger blocks, and it turns out that only wealthy miners can afford that. This creates the risk of centralization, hence questions of security.
SegWit is considered one of the most significant Bitcoin upgrades, but it hasn’t affected its scalability much. Even if absolutely all users sent SegWit transactions, Bitcoin’s TPS rate wouldn’t increase dramatically. The upgrade was a good try to improve the network, and it worked well in terms of transaction fees (SegWit payments cost less compared to legacy ones), but it hasn’t met the rising demand for Bitcoin’s scalability.
Ethereum’s sharding promises to be much more efficient in this regard. However, as the network’s complexity and the number of chains grow, their potential vulnerability has also increased — this forced Ethereum developers to do a lot of research and tests. That’s why the implementation of sharding stretched on for many years: it was first included in the roadmap in 2019, but now it’s clear that shard chains won’t go live before 2023.
There’s also a certain risk in consensus algorithms that focus on scaling while abandoning decentralization. The model of the EOS blockchain allows it to process a few thousand TPS, but there are only 21 nodes that have to reach a consensus to validate blocks. This is overly centralized — gaining control over at least 51% of nodes to attack the network is relatively easy.
Why Off-Chain Scaling Solutions?
On-chain scaling is a tough nut to crack, and each solution has limitations in terms of security and decentralization. This is why ways to scale networks without altering the blockchain itself have emerged and are referred to as off-chain solutions.
One of the reasons why they are gaining traction is the unwillingness of the crypto community to compromise when it comes to security. Bitcoin is the best example here — the values of the first cryptocurrency are scarcity, immutability, and resistance to censorship, and they are protected by a super decentralized network that is extremely difficult to hack. Such a renowned security level helped Bitcoin gain trust from actors like Tesla and Microstrategy. The community doesn’t want to change that, and Layer-2 scaling solutions for Bitcoin have started to emerge.
We’ve already discussed Ethereum and its course for sharding, but there are many off-chain solutions that scale the existing version of the Eth blockchain. The implementation of Proof of Stake and sharding takes years, while the demand for high Ethereum scalability is present right now. This sparked the development of a few solutions that we will review below.
Approaches to Off-Chain Scaling
Off-chain scaling means processing a part of transactions outside the blockchain or a few transactions packed into one. By doing so, we significantly increase the number of payments that the network can handle, hence we also improve their speed. High-value transactions that need strong security can still be verified in a traditional way, while other ones can take advantage of fast and cheap off-chain methods.
Here are a few examples of them:
- Lightning Network (Bitcoin) is a Layer-2 scaling solution for sending BTC almost instantly and at incredibly low fees. Here’s how it works: two parties create a payment channel via a smart contract and exchange Bitcoin through it. Once they decide to close the channel, all transactions that have been made are organized into one and are validated through a regular block mining procedure.
- Rollups is a technology that allows grouping a few Ethereum transactions and processing them as one using smart contracts. There are two kinds of rollups: Optimistic and ZK-Rollups. Arbitrum and Boba Network are among the top platforms using the technology.
- Sidechains are blockchains that run in parallel with the parent chain and connect with it via a two-way peg. For instance, Polygon is an Ethereum sidechain that leverages the main network’s security by processing, again, a bunch of transactions as a single one.
- Plasma chains are another scaling solution for Ethereum. They are smaller copies of the original Ethereum network that use Optimistic rollups. Although plasma chains allow for high throughput and low transaction fees, their functionality is relatively limited.
- State channels for Ethereum work much like the Lightning Network: they help two peers establish a connection and make as many payments as they want. When the channel closes, all transactions are grouped into one and processed in the Ethereum mainnet.
Summary and Prospects
There’s an abundance of ways to improve the scalability of a given cryptocurrency — the architecture of blockchain and the creativity of developers make that possible. On-chain solutions are mainly used in hard forked coins to raise the scalability of a predecessor, or, as in the case of Ethereum, the protocol is updated in a functioning cryptocurrency. Since making Ethereum more scalable could also make it less decentralized and secure, it took years to test all necessary features to make sure that the scaling roadmap is valid.
Off-chain methods are free from such risks, and they are extensively developed — mostly for older networks like Bitcoin and Ethereum that originally process very few transactions per second. Newer networks are inherently scalable since they were built at a time when it was already obvious that throughput should be high to satisfy users’ need to transact. However, maybe in a few years, cryptocurrency will witness a new round of adoption, and it will raise the demand for scaling even those networks that boast the best capacity today.