原文标题：《Modular Blockchain: A New Perspective on Functional Layer Controversies and DA Economics》
Original author: Zeke, YBB Capital
Original translation: Lucy, BlockBeats
The impossible triangle of blockchain has always been an obstacle that is difficult to cross in the industry. Many public chain projects have tried to overcome this gap through innovative architectural design in order to become the so-called "Ethereum killer". However, the reality is cruel. Over the years, Ethereum's dominance has remained solid, and the impossible triangle of blockchain has not been broken. So, is there a way for public chains to fill the gap of the impossible triangle? This is exactly the original intention of Mustafa Albasan to propose the concept of modular blockchain.
Modular Origins
The concept of modular blockchain originated from two white papers. The first one was co-authored by Mustafa Albasan and Vitalik in 2018, titled "Data Availability Sampling and Fraud Proofs". This paper explains how to solve the scalability problem of blockchain while maintaining security and decentralization. The specific method is to allow lightweight clients to receive and verify fraud proofs from full nodes, and design a data availability proof system to reduce the trade-off between on-chain capacity and security.
Then in 2019, Mustafa Albasan wrote a white paper on Lazy Ledger, detailing an innovative architecture in which the blockchain is only used to sort transaction data and ensure its availability, rather than to execute and verify transactions. This architecture aims to solve the scalability problems of existing blockchain systems. At the time, he called it a "smart contract client."
Smart contracts are executed through another execution layer on the client of Celestia (the first modular blockchain). Later, the emergence of Rollup made this concept more clear. Because the logic of Rollup is to execute smart contracts off-chain, aggregate the results into proofs, and then upload them to the execution layer of the "client".
Through in-depth thinking on blockchain architecture and new expansion technologies, he defined a new paradigm called "modular blockchain".
What is a modular blockchain?
Traditional monolithic blockchain architecture usually consists of the following four functional layers:
Execution layer: This layer is mainly responsible for processing transactions and executing smart contracts, including transaction verification, execution, and status updates.
Data availability layer: In a modular blockchain, the data availability layer ensures that data in the network can be accessed and verified. This layer usually includes functions such as data storage, transmission, and verification to ensure transparency and trust in the blockchain network.
Consensus layer: This layer is responsible for the agreement between nodes and achieves consistency of data and transactions in the network. Transactions are verified and new blocks are created through a specific consensus algorithm such as Proof of Work (PoW) or Proof of Stake (PoS).
Settlement layer: This layer is responsible for completing the final settlement of transactions, ensuring that the transfer and records of assets are permanently stored on the blockchain, and determining the final state of the blockchain.
Monolithic blockchains integrate these components into the same system. This highly integrated design often leads to inherent problems such as poor scalability, poor flexibility, and difficulty in maintenance and updating.
However, Celestia believes that a single blockchain no longer needs to take on all tasks alone. The future development of Web3 will be "modular blockchain", which will build better systems by modularizing the blockchain and dividing its processes into multiple "proprietary layers", each of which handles a specific functional layer. In addition, these systems should be independent, secure and scalable.
Modular design principles
If a system is designed into smaller parts that can be split, replaced, or substituted, then the design is modular. The core idea is to focus on doing a few specific things well (partial or single functional layer) instead of trying to cover everything. Cosmos Zones, Polkadot Parachain, etc. are all examples of modular projects that we are familiar with in the past.
new perspective
From a new perspective of modularity, the redesign space of monolithic blockchains and their related modular stacks will be greatly expanded. Various modular blockchains with different specific uses and architectures can be combined together to work together, and the diverse design possibilities have spawned many interesting and creative projects. Next, we will explore the current controversy about different functional layers and how Celestia interprets "modularity" from a modular perspective.
The execution layer is centered on Ethereum
If we consider Rollup as a modular execution layer, we will find that most modular execution layer projects are built on Ethereum. Obviously, this is because Ethereum has abundant resources as a moat, its degree of decentralization is the best choice, but its scalability is relatively poor, so there is great potential for redesign at the functional level.
By comparing the dismal performance of the recently launched Move system language public chains (APT, SUI) and the unprecedented craze for Layer2 on Ethereum, we can see that the narrative of blockchain infrastructure has shifted from developing public chains to developing Ethereum Layer2. So is modularity a good thing or a bad thing? Will the Ethereum-centric execution layer stifle public chain innovation?
Blockchain expansion view
First, from the perspective of the execution layer, the existing chains are reclassified. Here, we refer to Nosleepjon’s article “Tatooine’s Double Sun” to explain the current execution layer classification of blockchains.
Currently, blockchain can be divided into the following four categories:
Single-threaded monolithic blockchain: This type of blockchain processes only one transaction at a time. Due to its performance limitations, many projects have turned to Rollup or horizontal scaling solutions. Representative projects include: Ethereum, Polygon, Binance Chain, and Avalanche.
Parallel processing monolithic blockchain: This type of blockchain is able to process multiple transactions simultaneously. Representative projects include: Solana, Monad, Aptos, and Sui.
Single-threaded modular blockchain: This type of modular blockchain processes one transaction at a time. Representative projects include: Arbitrum, Optimism, zkSync, and Starknet.
Parallel processing modular blockchain: This type of modular blockchain can process multiple transactions at the same time. Representative projects include: Eclipse and Fuel.
Monolithic parallel processing architecture and modular architecture
There has been a lot of discussion about which approach to take, especially when it comes to the concepts of modularity vs global parallelism. There are three main camps of opinion:
Modular camp: Modular advocates (many of whom are also Ethereum supporters) believe that a single blockchain cannot solve the blockchain's impossible triangle problem. Stacking Lego blocks on Ethereum is considered the only way to achieve scalability while maintaining security and decentralization. Moreover, modularity has more control and customizability.
Monolithic parallel processing camp: This camp (quoting the views of Kodi and espresso in "Monolithic vs. Modular: Who is the future of blockchain?") believes that the new public chain architecture of monolithic parallel processing (such as Move system, Solona, etc.) has a high degree of integration and the overall performance will be better than the modular fragmented design. At the same time, the modular architecture is not safe, especially when a large amount of cross-chain communication is required, hackers have a wider attack surface.
Neutral camp: Of course, there are also people who hold a neutral attitude and believe that the two can eventually coexist. For example, Nosleepjon believes that the final game is that both sides have their own advantages, the competition between public chains will still exist, and Rollup will also compete with each other.
Summarize
The focus of this issue can actually be attributed to whether the frictional disadvantages of the modular solution (such as insufficient cross-chain security, poor system processes, etc.) outweigh the centralization issues of the new public chain. From the perspective of market debate, neither the shortcomings of the Rollup centralized isolator nor the security risks of the cross-chain bridge have led people to turn to new public chains. This is because all of these problems seem to have room for improvement, and the new public chain cannot replicate the huge ecological moat and decentralized advantages of the Ethereum chain.
On the other hand, although the new public chain has performance and integration advantages in architecture, its ecology is too similar to the Ethereum ecology, with a high degree of homogeneity and insufficient liquidity. Lacking a dedicated application that can reflect its own architectural advantages, there is naturally no reason for people to abandon the Ethereum ecology. The plasticity of Rollup is high enough, and there is still a lot of room for improvement of Rollup with new architecture in the future.
When Rollup also has most of the advantages of non-EVM chains, it will be difficult for "Solana Summer" to appear in the future. So in this case, I think the friction disadvantage of modular solutions is smaller than the problem of public chain centralization. And the neutral situation does not seem to exist. Ethereum's siphon effect will be like the "iPhone", attracting a large number of developers who focus on scalability to Layer2, and the new public chain will become a ghost town.
Therefore, regarding the future of infrastructure, I am undoubtedly more inclined to modularization. Ethereum's classified expansion will be the end of the public chain game, the Layer2 competition between general chains, and the Layer3 competition between super application chains.
This is also confirmed by the projects currently funded in the primary market. Apart from a large number of Ethereum Layer2 projects and Bitcoin expansion projects, there are almost no new public chains.
However, this industry has always been built on Ethereum, and the current trend seems to be too centralized. This situation is indeed worth thinking about. Lack of competition may hinder the development of an industry that needs diversity and more choices. If the user experience gradually becomes homogenized, how new public chains will create opportunities for breakthroughs is not yet clear. While Ethereum continues to improve its own defects, how to find larger gaps to accurately strike non-EVM systems needs to be focused on.
DA Program Competition
Recently, the industry has been discussing the shift from the execution layer to the data availability layer (DA layer), especially the question of which data availability solution Rollup should adopt. The discussion originated from a tweet by Dankrad Feist, a researcher at the Ethereum Foundation, and explored various aspects of the topic. In his opinion, Rollup without Ethereum DA does not belong to Layer2. Therefore, will the previous war on Layer1 evolve into a war between orthodox (with Ethereum DA) Layer2 and unorthodox Layer2? There are currently three main solutions for DA in the industry:
Public chain as settlement layer
Taking Ethereum as an example, the fees submitted to Ethereum when conducting transactions in Rollup mainly include the following categories:
Execution Fee: This is compensation for the computational resources required to execute the transaction. It includes the gas required to execute the transaction, which is usually proportional to the complexity and execution time of the transaction. In Rollup, the execution fee may include the cost of executing the transaction off-chain, as well as the cost of generating and verifying the transaction proof.
State Fees: State fees are associated with updating the state on the Ethereum mainchain. In Rollup, this includes the fee for submitting a new state root to the mainchain. Each time a Rollup aggregator generates a new state root and submits it to the mainchain, a state fee is incurred. This fee can be proportional to the frequency and complexity of state updates.
Data availability fee: The fee for publishing data to Layer1.
Among these fees, data availability fees account for the largest proportion and are relatively costly. For example, on May 6 this year, Arbitrum paid 376.8ETH GAS fees to Ethereum in a single day due to a surge in Ethereum GAS fees.
This is because Rollup uploads data to Ethereum in the form of Calldata uploads and stores this data permanently, which makes the cost very expensive. However, the security and legality of Rollup are the best of the three options, and the cost reduction of this option is currently waiting for the EIP-4844 update after the Cancun upgrade. By introducing the transaction format and using Blob to carry Transactions, the transaction format has one more Blob bit than the ordinary transaction format to carry Layer2 data. In addition, Blob data will be deleted by the node after 1 month, which greatly saves storage space.
Blob's transaction format provides cheaper data availability than Calldata. On the one hand, Calldata exists in the Execution Payload, while Blob data is stored in the Prysm node or Lighthouse node (not Geth), which consumes more resources when the contract needs to read Calldata. On the other hand, Blob data is short-term storage, and the node will delete Blob data after one month. Despite this, the GAS cost will still be higher than the latter two solutions.
Validiums DA Mode
For application chain type Rollups (such as dYdX, Immutable, etc. in the past), the Layer 2 scalability engine introduced by the header Rollup project is usually adopted (the most common one is StarkEx, but the Zk series header project also has similar plans). In the DA mode, due to the large amount of computation on the application chain, they prefer to use Validiums, which is a low-cost, high-throughput solution.
Validium aims to leverage off-chain data availability and computation, similar to ZK-rollup, by issuing zero-knowledge proofs to verify off-chain transactions on Ethereum. However, unlike ZK-rollup, which keeps data on-chain, Validiums keeps data off-chain, which costs 90% less than using Ethereum, making it the most cost-effective solution among the alternatives.
But since the data is still off-chain, the entity operator of Validium can freeze the user's funds. To prevent this extreme situation, the Data Availability Committee (DAC) plan needs to be introduced again. The DAC must confirm that the data has been received by signing each state update with its quorum. This is a controversial approach because you must first trust the security of the entity, not the chain itself. Dankrad Feist (creator of the above-mentioned EIP-4844) directly named this plan on Twitter.
Modular DA
From a modular perspective, there are many ways to redesign the DA layer, which may lead to differences in the specific implementation of each project. Therefore, it takes a lot of space to describe the modular DA project in detail. Among them, the Celestia project is used as a representative to explain the design of the DA project.
Celestia
As the first project to propose the concept of modular blockchain, Celestia has a high profile and pioneering position in the field. Its vision is to solve the problems of blockchain scalability and modularity. Built on the COSMOS architecture, Celestia provides developers with greater flexibility, allowing them to easily deploy and maintain blockchain applications. At the same time, by providing dApp creators and blockchain developers with a modular and scalable blockchain architecture, Celestia supports the needs of various applications and services, reducing the cost and complexity of deploying blockchains.
Working principle and architecture
Decoupled execution: The logic of Celestia is to decompose the protocol into different layers, each focusing on specific functions, which can be recombined to build blockchains and applications. Celestia focuses mainly on the consensus layer and data availability layer in the hierarchy. Similar to some Layer1s, Celestia uses Tendermint, a Byzantine Fault Tolerant (BFT) consensus algorithm, to sort transactions. But unlike other Layer1s, Celestia does not handle the validity of transactions or execute transactions. It only packages, sorts, and broadcasts transactions, and all transaction validity rules are enforced by the client's Rollup node (that is, the decoupling of the consensus layer and the execution layer is achieved).
A key point worth noting is "no reasoning about transaction validity". This means that malicious blocks carrying hidden transaction data can also be published to Celestia. So, how should the verification process be implemented? Celestia introduces two core technologies here: 2D Reed-Solomon encoding and data availability sampling (DAS).
The monolithic architecture of a single blockchain is in stark contrast to Celestia’s modular architecture.
DAS: This scheme allows light nodes to verify the availability of block data without downloading the entire block. Light nodes only need to sample a portion of the block data (the specific implementation relies on 2D Reed-Solomon encoding, see below for details). Unlike the previously mentioned Dac, DAS does not rely on the security of trusted entities; as long as the chain is sufficiently decentralized, the data can be trusted.
2D Reed-Solomon Coding (Erasure Coding): The core idea of 2D Reed-Solomon coding is to apply Reed-Solomon coding on rows and columns respectively. In this way, even if some rows and columns of the two-dimensional data are wrong, they can be corrected. By encoding the block data, the block data is divided into kk blocks, arranged into a kk matrix, and expanded into a 2k2k extended matrix through multiple Reed-Solomon encodings. 4k independent Merkle roots of the rows and columns of the extended matrix are calculated, and these Merkle roots are used as block data commitments in batches.
Celestia light nodes sample 2k2k data blocks. Each light node randomly selects a unique set of coordinates in the extended matrix and queries the full node for data blocks about these coordinates and the corresponding Merkle proofs. Each data block that receives a correct Merkle proof is broadcast to the network.
In abstract terms, the block data can be divided into square matrices (e.g. 8x8), and by encoding, additional "checksum" rows and columns are added to the original data to form a larger square matrix (e.g. 16x16). By randomly sampling part of the data in this large square matrix and verifying its accuracy, the integrity and availability of the overall data can be ensured. Even if part of the data is lost or damaged, the entire data can still be restored using the checksum data.
Block Scaling: Celestia implements features that scale with the number of light nodes. Celestia maintains security as long as there are enough nodes in the network to sample the entire block. This means that as more nodes join the network to sample, the block size can increase accordingly without sacrificing security or decentralization. However, on traditional monolithic blockchains, increasing the block size may sacrifice decentralization because larger block sizes increase the hardware requirements for nodes to download and verify data.
Sovereign Rollup: This is a concept first proposed by Celestia, which combines multiple blockchain design elements, including Layer1 blockchain, rollup, and Mastercoin in the early Bitcoin network. The main difference between sovereign Rollup and smart contract Rollup (such as Optimism, Arbitrum, zkSync, etc.) is the transaction verification method. In smart contract Rollup, transactions are verified by smart contracts deployed on Ethereum. In sovereign Rollup, the Rollup node itself is responsible for verifying transactions.
Sovereign Rollup publishes its transactions to other blockchains (such as Celestia) for ordering and data availability processing. Then, the nodes of the sovereign Rollup confirm the correct chain. This design enables sovereign Rollup to inherit multiple security properties from the DA layer, including liveness, security, reorganization resistance, and censorship resistance.
For smart contract Rollup, upgrades depend on the smart contract of the settlement layer. To upgrade Rollup, the smart contract needs to be modified. This may require multiple signatures to control who can initiate updates to the smart contract. Although it is common for the team to control the upgrade of multi-signatures, it is also feasible to control multi-signatures through governance. Because smart contracts are at the settlement layer, they are limited by the social consensus of the settlement layer.
Sovereign Rollup is upgraded through forks similar to Layer 1 blockchains. After a new software version is released, nodes can choose to update their software to the latest version. If the nodes do not agree to the upgrade, they can continue to use the old software. Such options allow people in the community who run nodes to decide whether to accept new changes. Even if most nodes are upgraded, they cannot be forced to accept the upgrade. This feature makes sovereign Rollup a truly "sovereign" Rollup.
The Quantum Gravitational Bridge (QGB) is a key component of the Celestia ecosystem. It acts as a bridge between Celestia and Ethereum (or other EVM L1 chains), enabling data and asset transfers between the two networks. By introducing the concept of Celestium (EVM L2 Rollup), Celestia is used to achieve data availability, while Ethereum is selected as the settlement layer.
This allows Celestium to leverage the best of both worlds: the scalability and data availability of Celestia, and the security and decentralization of Ethereum. Validators on Celestia can run QGB, enabling Celestium to provide strong data availability guarantees for block data at a fraction of the cost of Ethereum calldata.
QGB is a key part of Celestia's vision of a scalable, secure, and decentralized blockchain ecosystem. It promotes the interoperability required for the future of blockchain technology. Currently, the project is working on Zk QGB to further reduce the gas cost of verification.
DA Economics
Let’s talk about how much economic value DA will have in the future.
This hypothesis was proposed by Jon Charbonneau, a researcher at Delphi, based on Polygon Hermez's prediction that they will eventually only need 14 bytes per transaction in Danksharding. According to the above EIP-4844 specifications, at 1.3 MB/s, Layer2 can reach about 100,000 TPS, with an expected revenue of a staggering $30 billion.
Driven by such huge interests, the competition in the DA market will be extremely fierce in the future. In addition to the three major solutions, Stark's Layer3, zkPorter and some other modular DA projects will also join the battle. Therefore, from the existing Layer2 projects, general chains are more inclined to use Ethereum DA, while application chains and long-tail chains will become the main customers of "unorthodox DA". I personally believe that modular DA and the rapidly developing Layer3 will be the mainstream choice in the future.
Conclusion
Moving towards decentralization is still the mainstream concept of the industry. Modular blockchain is essentially an extension of Ethereum's values and an attempt to break the impossible triangle of blockchain. Although its design is diverse, this also leads to the complexity of construction. Since there are many modules to choose from in modular construction, and there are potential blind box risks between different modules, how to build a more stable modular system has become an issue that needs attention. On the other hand, driven by the modular trend, dozens of Layer2s will further reduce liquidity, and cross-chain communication and security will also become the focus of future development. Recently, the modularization of Bitcoin has also become a hot direction, and some of these solutions have certain feasibility and deserve moderate attention.
Original link
