Modular vs Monolithic Blockchain Design
It’s 1913, engineers have designed a completely new system to mass-produce the Ford Model T that would change how modern humans worked forever. The assembly line became the standard in the car manufacturing industry, with other industries rushing to introduce it to their own factories as well. Instead of having car manufacturing where everyone worked on everything, the main innovation from the assembly line came from specialization. Workers mastered one specific task within the assembly line and that was all they were responsible for.
So how successful was this specialization approach? Well, the assembly line cut the amount of time to make a Model T from 12.5 hours to 93 minutes. The number of cars made per day went from 100 to about 1000, a 10x growth. Specialization caused productivity and efficiency to skyrocket and even helped lower the prices for the Model T.
The Blockchain Trilemma
Blockchains still haven’t reached the chasm of specialization and hyper-efficiency needed to support mass adoption in a truly decentralized manner. They’re still in the pre-1913 era relative to the car factories. This is shown in The Blockchain Trilemma, which forces Blockchain to make trade-offs that prevent them from achieving all of the three following conditions:
Decentralized: A blockchain system that doesn’t rely on a central point, or a cartel of central points
Scalability: Being able to handle a large number of transactions
Secure: Being able to defend itself from bugs, attacks, and keeping the users’ funds safe.
This is why you should be suspicious of blockchains claiming to process thousands of transactions per second, as chances are that they made a trade-off somewhere to achieve it. A good example is the Delegated Proof of Stake (dPOS) consensus mechanism. This creates a small cartel of nodes that are processing transactions quickly and efficiently. The trade-off for this scaling solution is that it’s highly centralized and insecure as it’s pretty easy for an entity to take over the chain, or force the nodes to roll back transactions. This became a meme in 2017 as all of the “Ethereum killers” back then were centralized using dPOS to make transactions fast and cheap.
Its a trilemma, not a law
The Layer 1 smart contracts in 2017 were basically memes as they couldn’t solve the trilemma and used slick marketing to onboard users. 2021-2022 is a different story though. Ethereum, alongside of high-quality Layer 1’s are moving towards actually solving The Blockchain Dilemma. It’s important to note that the trilemma is just a model to show the main challenges that blockchains face in the past and present. There isn’t a rule stating that it’s impossible to achieve all three in the future.
Modular & Monolithic
Monolithic & The Components
The reason the dilemma exists is that blockchains are generally monolithic, including Ethereum in its current state and most other L1s. Just like the inefficiency of a car factory before 1913, monolithic blockchains try to inefficiently have all of the components of a blockchain under one system. The three main components are:
Consensus: The system that defines how blocks are added to the chain and how the nodes verify that the chain is correct and not corrupted.
Execution: The data, calculations, and transactions needed to update the blockchain to the next block.
Data Availability: The data that the node hosts. The amount of data available is called the block space and that data is appended forever to the blockchain.
The issue with monolithic blockchains is that since all three components are under the same layer, trying to optimize one component will cause tradeoffs and restrictions to the other components. Let’s look back at the dPOS example. By reducing the number of nodes needed to process transactions, the execution becomes much faster, while causing a drawback to the consensus by increasing centralization and fraud risk.
Modular
A modular blockchain is like an assembly line by taking the three blockchain components and separating them. By creating this separation and specialization, each component can be optimized to work more efficiently than under a monolithic system.
This concept of separation is what led to the car factory to produce 10x the amount of cars while making these cars much cheaper to buy. In blockchain terms, this would allow an exponential increase of transactions, while making these transaction fees much cheaper. The assembly line led to massive economic growth in the world. Modular structures will create an explosion of digital economic growth as the full potential of blockchains can be realized. This type of blockchain architecture is the closest to solving the blockchain dilemma by maximizing decentralization, scalability, and security without requiring any trade-offs.
Examples of modular chains
Ethereum: In its current state Ethereum is monolithic but the future is modular. L2 rollups can focus on the execution/ transactions component and disregard consensus and data availability as it carries the same security guarantees as to the main chain through a cryptographic proof. This will allow L2s to process exponentially more transactions at a cheaper cost than what the main Ethereum chain is capable of.
The second way that Ethereum plans to go modular is by sharding. Sharding essentially breaks the entire network into different portions called shards (groups of nodes) and would be independent of other shards. Sort of like acting as independent blockchains that are tied to Ethereum. Currently, every single node is responsible for every single transaction which creates inefficiencies. By splitting the nodes up based on which transactions they process, it should reduce congestion and continue to be decentralized as long as each shard has a certain amount of nodes.
Avalanche: The Avalanche architecture is slightly similar to sharding, but instead their model is based on subnets. These subnets are a group of validator nodes that coordinate in order to come up to a consensus but without requiring every node to process every transaction. They’re split just like in the sharding model. The three built-in blockchains within Avalanche are the X-Chain, P-Chain, and the C-Chain and each has specific functions within the Avalanche ecosystem that specialize in a certain aspect of the overall blockchain.
Cosmos: The best way to describe Cosmos is that it’s a massive toolbox that allows anyone to create modular blockchains that can communicate with each other (interoperable). The interoperability is done by connecting to the Inter-Blockchain Communication (IBC) protocol which allows these blockchains to transfer data and assets with each other. The Cosmos Hub, which is the native blockchain in the ecosystem, is responsible for governance, voting, and staking of their ATOM token.
There are other modular chains, but I felt that these can be considered the top 3.
(I will be writing in-depth articles about L2s, Avalanche, and Cosmos in the future as this only scratches the surface)
-Captain BTC