In my last post, we dove into blockchains and explored opportunities to leverage them in ways which can fundamentally disrupt our lives. From transacting payments over intermediary-less networks, to cutting the middleman from content creation through NFTs, blockchains can bring transparency, accountability and security to traditionally “black box” industries. Blockchains have the potential to reset how we perceive computer networks and reimagine the economics of business models.
Get onchain to be unchained
Blockchains are creating networks which are more equitable and bringing back power to the customer or creator. But to have meaningful and long-lasting impact, blockchains need to benefit from network effects. Networks are like parties; the more the merrier. So, more users have to be “onchain” (i.e., transacting on blockchains) for these networks to be successful. As blockchain networks grow, the value of these networks grow as well (Greyscale 2023).
However, recent statistics indicate that there are over 425 million users of cryptocurrency, which is one of the biggest derivatives of blockchain technology, and possibly, the closest number we have to understanding how many users there are of blockchains. But that only makes up 5.2% of the global population and 7.9% of people who have access to the internet. That’s ridiculously small!
The ramification of this is that the vast majority of people only have access to today’s internet that is largely controlled by big corporations which have created oligopolies and centralized the internet in a way that deters competition while extorting high fees. As Chris Dixon, GP at a16z, puts it in his book Read Write Own: the internet has gotten “intermediated” over time, going from “permissionless to permissioned”.
Scaling blockchain networks is a challenge
Theoretically, adding more nodes to the network should help balance the higher consumption of resources, as new users are onchain. However, the costs associated with validating transactions are high. A Bitcoin miner needs specialized, high performance hardware, such as an application-specific integrated circuit (or ASIC) as well as heating/cooling systems etc, which costs upwards of $20,000. Meanwhile, on Ethereum, you need to stake 32 ETH (Ethereum’s native token) or ~$109,000 to be a validator node. [Now, there are ways around managing the staking requirement i.e. using staking pools, which pools funds, or accessing other forms of liquidity but there are downsides such as receiving partial benefits of a full validator node and loss of independence when centralized forces make decisions for staking pools.] A cost is incurred, one way or another. So if becoming a mining or validator node is not fully accessible to everyone, how does the blockchain network scale to allow for better reach?
Over the last few years, more specialized approaches within the blockchain architecture have emerged to help with scaling. The OG blockchains, i.e. Bitcoin and Ethereum, started with the “all-in-one” approach where all functions occur on the same blockchain. More recently, separate layers of specialization have been introduced, with “layer 2” solutions, which overlay the base blockchain, to the recent introduction of explicitly modular blockchains. Let’s take a closer look at the different functions of a blockchain, and how the architecture has transitioned from “generalist” to “specialist” and perhaps, going forward, may even loop back.
It starts from the core
A blockchain is a database or computer which is immutable and stores information across a distributed network. This involves four core functions:
- Execution: involves processing transactions provided by users and smart contracts for validation.
- Consensus: allows all nodes of a decentralized blockchain network to agree on the state of the virtual computer and the order of transactions.
- Settlement: secures the destination of transactions and resolves disputes, which deems them to be “final” or “settled”.
- Data availability: allows transaction data to be available, so that the state of the network can be verified. This function is required for #1, #2 and #3. This facilitates transparency in the blockchain.
Monolithic blockchains house all four functions mentioned above in a single layer. With all functions competing for the same resources, blockchain networks often cap the number of validator nodes since computing requirements increase with the expansion of the network. Even today, Bitcoin miners and Ethereum nodes require high-compute specifications in their hardware which becomes expensive and difficult for scaling of the networks. Given these high computing requirements, the speed of the networks also takes a hit. Bitcoin and Ethereum have lower transaction speeds compared to some centralized, Web2 systems, such as PayPal and Visa. While newer base blockchains or L1s have higher transaction speeds, such as Polygon and Solana, the higher speed often comes at the expense of other essential properties of a blockchain, such as security and decentralization. This tradeoff is often referred to as the “Blockchain Trilemma” (coined by Vitalik Buterin, the co-founder of Ethereum).
This paradigm shows that technology has not been able to successfully incorporate all three pillars of a blockchain in equal measure. For instance, if a blockchain has achieved decentralization and security, aka Bitcoin, it has not been able to achieve scale. If blockchains can’t handle a large number of nodes and scale, i.e. process and validate transactions, they can’t compete with conventional, corporate networks.
Evolving blockchains for global scale
To address the massive scaling needs to enable mass adoption of blockchains, innovative solutions are emerging, creating various layers in the blockchain network stack:
Sharding: typically at the base blockchain layer or L1s, this breaks up a single blockchain into multiple blockchains, called shards, and reduces the data burden on individual nodes. This improves the blockchain throughput and increases transaction speeds. There are some downsides. The threat of decentralization is bigger, since nodes are split between different blockchains, the network is left open to hostile threats, such as validator hacks. Less of a problem for large networks like Ethereum, smaller networks would bear the brunt of this. Data inconsistency is another issue, since discrepancies in data recorded could lead to different “states” being recorded in each shard. This hampers one of the core value propositions of blockchains. Blockchains such as Polkadot and NEAR have successfully integrated sharding into their networks to help with scaling. As part of Ethereum’s Dencun upgrade in March 2024, proto-danksharding was introduced (which provides temporary data storage mechanisms to help with data availability) which has already helped reduce transaction fees. This is expected to be an intermediate measure in the process of introducing sharding to the Ethereum mainnet.
Layer-2s (L2s): these are built on top of the L1s with the aim of improving the scalability of the blockchain. On Ethereum, roll-ups are a type of L2 which use smart contracts, or self executable code, which processes transactions off the main network, i.e. offchain, in batches and periodically records that information on the blockchain. This allows more transactions to be processed, which reduces transaction fees and improves processing speeds. On Bitcoin, this has been introduced through the Lightning network, which creates secure “state channels” with specific people, using smart contracts.
Modular: introduces the concept of separating functions into different layers or modules. In a modular paradigm, separating core functions into different blockchains allows more resources to be directed to an individual specialization which allows for faster processing. Celestia, for example, is a modular blockchain which focuses on data availability and consensus. This means its only goal is to provide consensus over the order of transactions and verify that the data is available. Execution and settlement occurs separately, while inheriting security benefits from Celestia.
Blockchains are at the cusp of mass adoption, but to achieve this, scalability is key. And the scaling measures which have been introduced certainly work towards this goal. With the introduction of Celestia and other data availability modular blockchains, L3s and even L4s have come about, taking modularity to a whole different level. I, for one, will be keeping a close eye on how these blockchains work through kinks related to security and data inconsistency, among others, in order to achieve scale.