Modular blockchain is an innovative blockchain design paradigm aimed at enhancing system efficiency and scalability through specialization and division of labor.
Figure 1: Modular blockchain example
Before the emergence of modular blockchain, a single monolithic chain was responsible for handling all tasks, including the execution layer, data availability layer, consensus layer, and settlement layer. Modular blockchain views these tasks as freely combinable modules to address these issues, with each module focusing on specific functions.
Execution Layer: Responsible for processing and validating all transactions, as well as managing blockchain state changes.
Consensus Layer: Achieves agreement on transaction order.
Settlement Layer: Used for completing transactions, verifying proofs, and bridging between different execution layers.
Data Availability Layer: Ensures that all necessary data is accessible to participants in the network for verification.
The trend of modular blockchain is not just a technological shift but also a crucial strategy to drive the entire blockchain ecosystem toward future challenges. GeekCartel will analyze the concept of modular blockchain and related projects, aiming to provide a comprehensive, practical interpretation of modular blockchain knowledge to help readers better understand modular blockchain and anticipate future development trends. Note: The content of this article does not constitute investment advice.
In 2018, Mustafa Albasan and Vitalik Buterin published a groundbreaking article that provided a new approach to addressing the scalability issues of blockchain. “Data Availability Sampling and Fraud Proofs“ introduced a method through which blockchain could automatically expand its storage space as the network nodes increased. In 2019, Mustafa Albasan further researched and authored “Lazy Ledger,” proposing a blockchain system concept that only deals with data availability.
Based on these concepts, Celestia emerged as the first Data Availability (DA) network adopting a modular structure. Built using CometBFT and Cosmos SDK, it is a Proof of Stake (PoS) blockchain that effectively improves scalability while maintaining decentralization.
The DA layer is crucial for the security of any blockchain as it ensures that anyone can inspect the transaction ledger and verify it. If a block producer proposes a block without all data available, the block can achieve final determinism but may contain invalid transactions. Even if the block is valid, data that cannot be fully verified will adversely affect the functionality of users and the network.
Celestia implements two key functions: Data Availability Sampling (DAS) and Namespace Merkle Trees (NMT). DAS allows light nodes to verify data availability without downloading the entire block. NMTs enable block data to be partitioned into separate namespaces for different applications, meaning applications only need to download and process data relevant to them, greatly reducing data processing requirements. Importantly, DAS allows Celestia to scale with the increasing number of users (light nodes) without compromising the security of end-users.
Modular blockchain is making it possible to build new chains in unprecedented ways, where different types of modular blockchains can collaborate in different architectures and for different purposes. Celestia’s official proposals for modular architecture designs and examples demonstrate the flexibility and composability of modular blockchains.
Figure 2: Layer1 and Layer2 architecture
Layer 1 and Layer 2: Celestia refers to this as naive modularization, initially built for the scalability of Ethereum as a monolithic Layer 1, with Layer 2 focusing on execution while Layer 1 provides other key functionalities.
Figure 3: Execution, settlement and DA layer architecture
Execution, settlement, and data availability: Optimized modular blockchains can decouple the execution, settlement, and data availability layers across specialized modular blockchains.
Figure 4: Execution and DA layer architecture
Execution and DA: Since the purpose of implementing modular blockchain is flexibility, the execution layer is not limited to simply publishing its blocks to the settlement layer. For example, a modular stack can be created that does not involve the settlement layer, only the execution layer above the consensus and data availability layers.
In this modular stack, the execution layer would be sovereign, publishing its transactions to another blockchain, typically used for ordering and data availability but handling its own settlement. In the context of the modular stack, sovereign Rollup is responsible for execution and settlement, while the DA layer handles consensus and data availability.
The difference between sovereign Rollup and smart contract Rollup is:
Currently, Rollkit and Sovereign SDK provide frameworks for deploying sovereign Rollup testnets on Celestia.
Before introducing modularization of the execution layer, we should understand what Rollup technology is.
Currently, modularization technology of the execution layer relies heavily on Rollup, which is a scaling solution that operates off-chain from Layer 1. This solution executes transactions off-chain, meaning it occupies less block space and is one of Ethereum’s important scaling solutions. After executing transactions, it sends a batch of transaction data or execution proofs to Layer 1 for settlement. Rollup technology provides a scalability solution for Layer 1 networks while maintaining decentralization and security.
Figure 5: Rollup technical architecture
Taking Ethereum as an example, Rollup technology can further enhance performance and privacy by using ZK-Rollup or Optimistic Rollup.
Initially, Ethereum adopted sidechains and sharding technology for scalability, but sidechains sacrificed some decentralization and security to achieve high throughput. The development of Layer 2 Rollups has progressed much faster than expected and has already provided significant scalability, with even more to come after the implementation of Proto-Danksharding. This means that there is no longer a need for “shard chains,” which have been removed from Ethereum’s roadmap.
Ethereum outsources the execution layer to Layer 2s based on Rollup technology to relieve the burden on the main chain, and the EVM provides a standardized and secure execution environment for smart contracts executed on the Rollup layer. Some Rollup solutions are designed with compatibility with the EVM in mind, allowing smart contracts executed on the Rollup layer to still leverage the features and functionalities of the EVM, such as OP Mainnet, Arbitrum One, and Polygon zkEVM.
Figure 6: Ethereum’s layer 2 scaling solution
These Layer 2s execute smart contracts and process transactions but still rely on Ethereum for the following operations:
Settlement: All Rollup transactions are settled on the Ethereum mainnet. Users of Optimistic Rollups must wait for the challenge period to pass or for transactions to be deemed valid after fraud proofs are computed. Users of ZK Rollups must wait for validity to be proven.
Consensus and Data Availability: Rollups publish transaction data to the Ethereum mainnet in the form of CallData, allowing anyone to execute Rollup transactions and reconstruct their state if necessary. Before confirmation on the Ethereum main chain, Optimistic Rollups require significant block space and a 7-day challenge period. ZK Rollups provide instant finality and store data available for verification for 30 days, but require significant computational power to create proofs.
B² Network is the first ZK-Rollup on Bitcoin, enabling increased transaction speeds without sacrificing security. Leveraging Rollup technology, B² Network provides a platform for running Turing-complete smart contracts for off-chain transactions, thereby improving transaction efficiency and minimizing costs.
Figure 7: B² Network architecture
As shown in the diagram, the ZK-Rollup layer of B² Network adopts the zkEVM solution, responsible for executing user transactions within the Layer 2 network and producing relevant proofs.
Unlike other Rollups, the B² Network ZK-Rollup consists of multiple components, including the Account Abstraction Module, RPC Service, Mempool, Sequencers, zkEVM, Aggregators, Synchronizers, and Prover. The Account Abstraction Module implements native account abstraction, allowing users to programmatically incorporate higher security and better user experience into their accounts. zkEVM is compatible with EVM and can also assist developers in migrating DApps from other EVM-compatible chains to B² Network.
Synchronizers ensure that information is synchronized from B² nodes to the Rollup layer, including sequence information, Bitcoin transaction data, and other details. B² nodes act as off-chain validators and executors of several unique functions within the B² Network. The Bitcoin Committer module in B² nodes constructs a data structure to record B² Rollup data and generates a Tapscript called “B² ciphertext.” Then, the Bitcoin Committer sends a UTXO of one satoshi to a Taproot address containing the $B^{2}$ ciphertext, and the Rollup data is written into Bitcoin.
Additionally, the Bitcoin Committer sets a time-locked challenge, allowing challengers to dispute the commitment of zk proofs. If there are no challengers during the time lock or if the challenge fails, the Rollup is eventually confirmed on Bitcoin; if the challenge is successful, the Rollup is rolled back.
Whether it’s Ethereum or Bitcoin, Layer 1 is essentially a single chain that receives extended data from Layer 2. In most cases, the capacity of Layer 2 also depends on the capacity of Layer 1. Therefore, the implementation of the Layer 1 and Layer 2 stack is not ideal for scalability. When Layer 1 reaches its throughput limit, Layer 2 is also affected, which may lead to increased transaction fees and longer confirmation times, affecting the efficiency and user experience of the entire system.
In addition to Celestia’s DA solution being favored by Layer 2s, other innovative solutions focused on DA have emerged, playing crucial roles in the entire blockchain ecosystem.
EigenDA is a secure, high-throughput, and decentralized DA service inspired by Danksharding. Rollup can publish data to EigenDA to achieve lower transaction costs, higher transaction throughput, and secure composability throughout the EigenLayer ecosystem.
When building decentralized temporary data storage for Ethereum Rollup, data storage can be directly handled by EigenDA operators. Operators participate in network operation, responsible for processing, verifying, and storing data, and EigenDA can horizontally scale with the increase in staking and operators.
EigenDA combines Rollup technology while moving the DA part off-chain for scalability. As a result, actual transaction data no longer needs to be replicated and stored on every node, reducing bandwidth and storage demands. On-chain, only metadata related to data availability and accountability mechanisms are processed (accountability ensures data is stored off-chain and can be verified for integrity and authenticity when necessary).
Figure 8: Basic data flow of EigenDA
As shown in the diagram, Rollup writes transaction batches to the DA layer. Unlike systems that use fraud proofs to detect malicious data, EigenDA divides the data into blocks and generates KZG commitments and multi-reveal proofs. EigenDA requires nodes to download only a small amount of data [O(1/n)], rather than downloading the entire blob. The fraud arbitration protocol of Rollup can also verify whether the blob data matches the KZG commitments provided in the EigenDA proof. Through this verification, Layer 2 chains can ensure that transaction data of the Rollup state root is not manipulated by sequencers/proposers.
Nubit is an scalable, Bitcoin-native DA layer aiming to advance the future of Bitcoin by enhancing data throughput and availability services to meet the growing demands of the ecosystem. Their vision is to integrate the vast developer community into the Bitcoin ecosystem, providing them with scalable, secure, and decentralized tools.
The team behind Nubit consists of professors and PhD students from UCSB (University of California, Santa Barbara), with outstanding academic reputations and global influence. They are not only proficient in academic research but also have rich experience in implementing blockchain engineering. The team, along with domo (the creator of Brc20), co-authored a paper on modular indexers, incorporating the design of the DA layer into the indexer structure of the Bitcoin meta protocol, contributing to the establishment and formulation of industry standards.
Nubit’s core innovations lie in its consensus mechanism, trustless bridging, and data availability, utilizing innovative consensus algorithms and the Lightning Network to inherit Bitcoin’s fully censorship-resistant characteristics and improve efficiency through DAS:
Consensus Mechanism: Nubit explores an efficient consensus based on PBFT (Practical Byzantine Fault Tolerance) supported by SNARKs for signature aggregation. The combination of PBFT with zkSNARK technology significantly reduces the communication complexity of verifying signatures between validators, verifying the correctness of transactions without accessing the entire dataset.
DAS: Nubit’s DAS is achieved by multiple rounds of random sampling of small portions of block data. Each successful sampling round increases the probability of complete data availability. Once the predetermined confidence level is reached, the block data is considered accessible.
Trustless Bridge: Nubit utilizes a Trustless Bridge leveraging the Lightning Network‘s payment channels. This approach not only aligns with local Bitcoin payment methods but also does not impose additional trust requirements. Compared to existing bridging solutions, it brings lower risks to users.
Figure 9: Basic components of Nubit
Let’s further review the complete system lifecycle as depicted in Figure 8 using a specific use case. Suppose Alice wants to complete a transaction using Nubit’s DA service (Nubit supports various data types, including but not limited to ciphertext, Rollup data, etc.).
The light client undergoes the DAS process to verify data availability. Additionally, after proposing a threshold number of blocks, checkpoints of this history are recorded on the Bitcoin blockchain through Bitcoin timestamps. This ensures that the validator set can thwart potential remote attacks and support rapid unbinding.
In addition to focusing on chains with specific modular layers, decentralized storage services can provide long-term support for the DA layer. There are also some protocols and chains that offer developers customized and full-stack solutions, allowing users to easily build their own chains without even needing to write code.
EthStorage is the first modular Layer 2 to achieve dynamic decentralized storage, offering DA-driven programmable key-value (KV) storage. It @ld-capital/%E4%BB%8Eethstorage-%E5%9B%9E%E7%9C%8B%E8%A2%AB%E5%B8%82%E5%9C%BA-%E5%86%B7%E8%90%BD-%E7%9A%84%E5%8E%BB%E4%B8%AD%E5%BF%83%E5%8C%96%E5%AD%98%E5%82%A8%E8%B5%9B%E9%81%93-d0a003220362">extends programmable storage to hundreds of TB or even PB at 1/100 to 1/1000 of the cost. EthStorage provides a long-term DA solution for Rollups and opens new possibilities for fully on-chain applications in gaming, social networks, AI, and more.
Figure 10: Application scenarios of EthStorage
Qi Zhou, the founder of EthStorage, has been fully dedicated to the Web3 industry since 2018. He holds a Ph.D. from the Georgia Institute of Technology and has previously worked as an engineer at top companies such as Google and Facebook. His team has also received support from the Ethereum Foundation.
As one of the core features of the Ethereum Cancun upgrade, EIP-4844 (also known as Proto-dank sharding) introduces temporary data blocks (blobs) for Layer 2 Rollup storage, enhancing the network’s scalability and security. The network does not need to verify every transaction in the block, only to confirm whether the blob attached to the block carries the correct data, significantly reducing the cost of Rollups. However, blob data is only temporarily available, meaning it will be discarded within a few weeks. This has a significant impact: Layer 2 cannot unconditionally derive the latest state from Layer 1. If a piece of data can no longer be retrieved from Layer 1, it may be impossible to sync the chain via Rollup.
With EthStorage as a long-term DA storage solution, Layer 2s can access complete data from its DA layer at any time.
Technical features:
EthStorage enables decentralized dynamic storage: Existing decentralized storage solutions can support large data uploads but cannot modify or delete them, only re-upload new data. EthStorage significantly enhances data management flexibility by achieving CRUD functionality (Create, Read, Update, Delete) through an original key-value storage paradigm.
Layer 2 decentralized solutions based on the DA layer: EthStorage is a modular storage layer that can run on any blockchain with EVM and DA to reduce storage costs (though many Layer 1s currently lack a DA layer), and it can even run on Layer 2.
High integration with ETH: The EthStorage client is a superset of the Ethereum client Geth, meaning that when running an EthStorage node, it can still participate in any Ethereum processes. A node can be both an Ethereum validator node and an EthStorage data node simultaneously.
EthStorage’s workflow:
AltLayer offers a versatile, no-code Rollups-as-a-Service (RaaS) solution. Designed for a multi-chain and multi-virtual machine world, the RaaS product supports both EVM and WASM. It also supports various Rollup SDKs, such as OP Stack, Arbitrum Orbit, Polygon zkEVM, ZKSync’s ZKStack, and Starkware, as well as different shared sequencing services (e.g., Espresso and Radius) and different DA layers (e.g., Celestia and EigenLayer), along with many other modular services for different layers of the Rollup stack.
With AltLayer, a multifunctional Rollup stack can be achieved. For instance, a Rollup designed for an application can be built using Arbitrum Orbit, with Arbitrum One as the DA and settlement layer. Meanwhile, a general-purpose Rollup can be constructed using the ZK Stack, with Celestia as the DA layer and Ethereum as the settlement layer.
Note: You might wonder why the settlement layer can be implemented by OP and Arbitrum. In fact, these Layer2 Rollup stacks are currently achieving interchain connectivity similar to what Cosmos proposed. OP has introduced the Superchain, where the OP Stack acts as the standardized development stack supporting Optimism technology, integrating different Layer2 networks and promoting interoperability among them. Arbitrum has proposed the Orbitchain strategy, which allows Layer3s, also known as application chains, to be created and deployed on the Arbitrum mainnet based on the Arbitrum Nitro tech stack. Orbit Chains can settle directly on Layer2s or directly on Ethereum.
Dymension is a modular blockchain network based on the Cosmos SDK, designed to ensure the security and interoperability of RollApps using the IBC standard. Dymension divides blockchain functionalities into multiple layers, with the Dymension Hub serving as the settlement and consensus layer, providing security, interoperability, and liquidity for RollApps, which act as the execution layer. The data availability (DA) layer is supported by Dymension protocol’s DA providers, allowing developers to choose the appropriate DA provider based on their needs.
The settlement layer (Dymension Hub) maintains a registry of RollApps and their critical information, such as state, sequencer lists, currently active sequencers, and execution module checksums. Rollup service logic is fixed within the settlement layer, forming a native interoperability hub. The Dymension Hub, as the settlement layer, has the following features:
In the previous section, we delved into modular blockchain systems and numerous representative projects. Now, we will shift our focus to comparative analysis among different ecosystems, aiming for an objective and comprehensive understanding of modular blockchain.
As we’ve seen, blockchain ecosystems are evolving towards modularity. In the past, blockchain networks operated in isolation, competing with each other, making it difficult for users, developers, and assets to move between different chains, thus restricting the overall development and innovation of the ecosystem. In the Web3 world, identifying and solving problems is a collaborative process. Initially, Bitcoin and Ethereum attracted significant attention as single chains, but as the limitations of single chains became apparent, modular chains began to gain attention. Therefore, the emergence of modular chains is not accidental but rather an inevitable development.
Modular blockchain enhances the flexibility and efficiency of chains by allowing individual components to be independently optimized and customized. However, this architecture also faces challenges such as increased communication latency and complexity in system interactions. In practice, the long-term benefits of modular architecture, such as improved maintainability, reusability, and flexibility, often outweigh its short-term performance losses. In the future, as technology advances, these issues will find better solutions.
GeekCartel believes that blockchain ecosystems have a responsibility to provide reliable foundational layers and common tools throughout the entire modular stack to facilitate seamless connections between chains. If ecosystems can be more harmonious and interconnected, users will be able to more easily use blockchain technology, attracting more new users to Web3.
Currently, there are several Restaking protocols emerging, effectively aggregating dispersed security resources through a re-staking mechanism to enhance the overall security of blockchain networks. This process not only addresses the issue of fragmented security resources but also strengthens the network’s defense against potential attacks while providing additional incentives for participants, encouraging more users to engage in network security maintenance. In this way, Restaking protocols open up new avenues for improving network security and efficiency, powerfully promoting the healthy development of blockchain ecosystems.
EigenLayer is a protocol built on Ethereum that introduces the Restaking mechanism, which is a new primitive for cryptographic economic security. This primitive allows the reuse of ETH at the consensus layer, aggregating ETH security across all modules, thus enhancing the security of DApps that rely on these modules. Users who stake native ETH or use Liquidity Staking Tokens (LST) to stake ETH can choose to join the EigenLayer smart contract to restake their ETH or LST, extending cryptographic economic security to other applications on the network and earning additional rewards.
As Ethereum moves towards a Rollup-centric roadmap, applications built on Ethereum are experiencing significant scalability. However, any module that cannot be deployed or proven on the Ethereum Virtual Machine (EVM) cannot absorb the collective trust of Ethereum. Such modules involve processing inputs from outside Ethereum, making their processing unverifiable within Ethereum’s internal protocols. These modules include sidechains based on new consensus protocols, data availability layers, new virtual machines, oracle networks, bridges, and more. Typically, such modules require an @GenesisLRT/what-are-avss-and-operators-00e1c51dab1c">Autonomous Verification System (AVS) with its own distributed verification semantics for validation. Usually, these AVSs are protected by their native tokens or have permissioned properties.
Currently, there are some issues within the AVS ecosystem:
Figure 11: Comparison between current AVS service and EigenLayer
In the architecture of EigenLayer, AVS (Autonomous Verification System) is a service built on the EigenLayer protocol, leveraging the shared security of Ethereum. EigenLayer introduces two novel approaches, through staking and free market governance, to achieve centralized security. These approaches help extend Ethereum’s security to any system and eliminate the inefficiency of existing rigid governance structures:
By combining these approaches, EigenLayer acts as an open market where AVSs can leverage the pooled security provided by Ethereum validators, encouraging validators to make more optimized trade-offs between security and performance through reward incentives and penalties.
Babylon is a Layer1 blockchain founded by Stanford University professor David Tse. The team consists of researchers from Stanford University, experienced developers, and business advisors. Babylon introduces the Bitcoin staking protocol, designed as a modular plugin to be used on various PoS (Proof of Stake) consensus algorithms, providing a primitive for re-staking.
Babylon leverages three aspects of Bitcoin—timestamping service, block space, and asset value—to propagate Bitcoin’s security to many PoS chains (such as Cosmos, Binance Smart Chain, Polkadot, Polygon, and other blockchains with robust and interoperable ecosystems), creating a stronger and more unified ecosystem.
Bitcoin timestamping resolves PoS long-range attacks:
Long-range attacks involve validators in a PoS chain unstaking and going back to a historical block where they were still validators, potentially starting a fork. This problem is inherent in PoS systems and cannot be completely solved solely by improving the consensus mechanism of PoS chains themselves. Both Ethereum and Cosmos, among other PoS chains, face this challenge.
After introducing Bitcoin timestamping, on-chain data of PoS chains will be stored on the Bitcoin chain with Bitcoin timestamps. Even if someone tries to create a fork of a PoS chain, the corresponding Bitcoin timestamp will definitely be later than the original chain, rendering the long-range attack ineffective.
Bitcoin staking protocol:
This protocol allows Bitcoin holders to stake their idle Bitcoin to enhance the security of PoS chains and earn rewards in the process.
The core infrastructure of the Bitcoin staking protocol is the Control Plane between Bitcoin and PoS chains, as illustrated in the diagram below.
Figure 12: System architecture with Control plane and Data plane
The Control Plane is implemented in the form of a chain to ensure it is decentralized, secure, censorship-resistant, and scalable. This control plane is responsible for various critical functions, including:
• Providing Bitcoin timestamping services for PoS chains to synchronize with the Bitcoin network.
• Acting as a marketplace, matching Bitcoin staking with PoS chains and tracking staking and validation information, such as the registration and refresh of EOTS (Epoch Time Oracle Service) keys.
• Recording the finality signatures of PoS chains.
By staking their BTC, users can provide validation services for PoS chains, DA layers, oracles, AVSs (Autonomous Verification Systems), etc. Additionally, Babylon can now offer services for Altlayer, Nubit, and other platforms.
Figures:
Articles:
In this emerging infrastructure paradigm, there is still much research and work to be done, and there are many areas that this article did not cover. If you are interested in any related research topics, please contact Chloe.
A big thanks to Severus and Jiayi for their insightful comments and feedback on this article.
This article is reproduced from [medium], the original title is “Modular Blockchain: The Last Piece of the Web3 Puzzle”, the copyright belongs to the original author [GeekCartel], if you have any objection to the reprint, please contact Gate Learn Team, the team will handle it as soon as possible according to relevant procedures.
Disclaimer: The views and opinions expressed in this article represent only the author’s personal views and do not constitute any investment advice.
Other language versions of the article are translated by the Gate Learn team, not mentioned in Gate.io, the translated article may not be reproduced, distributed or plagiarized.
Modular blockchain is an innovative blockchain design paradigm aimed at enhancing system efficiency and scalability through specialization and division of labor.
Figure 1: Modular blockchain example
Before the emergence of modular blockchain, a single monolithic chain was responsible for handling all tasks, including the execution layer, data availability layer, consensus layer, and settlement layer. Modular blockchain views these tasks as freely combinable modules to address these issues, with each module focusing on specific functions.
Execution Layer: Responsible for processing and validating all transactions, as well as managing blockchain state changes.
Consensus Layer: Achieves agreement on transaction order.
Settlement Layer: Used for completing transactions, verifying proofs, and bridging between different execution layers.
Data Availability Layer: Ensures that all necessary data is accessible to participants in the network for verification.
The trend of modular blockchain is not just a technological shift but also a crucial strategy to drive the entire blockchain ecosystem toward future challenges. GeekCartel will analyze the concept of modular blockchain and related projects, aiming to provide a comprehensive, practical interpretation of modular blockchain knowledge to help readers better understand modular blockchain and anticipate future development trends. Note: The content of this article does not constitute investment advice.
In 2018, Mustafa Albasan and Vitalik Buterin published a groundbreaking article that provided a new approach to addressing the scalability issues of blockchain. “Data Availability Sampling and Fraud Proofs“ introduced a method through which blockchain could automatically expand its storage space as the network nodes increased. In 2019, Mustafa Albasan further researched and authored “Lazy Ledger,” proposing a blockchain system concept that only deals with data availability.
Based on these concepts, Celestia emerged as the first Data Availability (DA) network adopting a modular structure. Built using CometBFT and Cosmos SDK, it is a Proof of Stake (PoS) blockchain that effectively improves scalability while maintaining decentralization.
The DA layer is crucial for the security of any blockchain as it ensures that anyone can inspect the transaction ledger and verify it. If a block producer proposes a block without all data available, the block can achieve final determinism but may contain invalid transactions. Even if the block is valid, data that cannot be fully verified will adversely affect the functionality of users and the network.
Celestia implements two key functions: Data Availability Sampling (DAS) and Namespace Merkle Trees (NMT). DAS allows light nodes to verify data availability without downloading the entire block. NMTs enable block data to be partitioned into separate namespaces for different applications, meaning applications only need to download and process data relevant to them, greatly reducing data processing requirements. Importantly, DAS allows Celestia to scale with the increasing number of users (light nodes) without compromising the security of end-users.
Modular blockchain is making it possible to build new chains in unprecedented ways, where different types of modular blockchains can collaborate in different architectures and for different purposes. Celestia’s official proposals for modular architecture designs and examples demonstrate the flexibility and composability of modular blockchains.
Figure 2: Layer1 and Layer2 architecture
Layer 1 and Layer 2: Celestia refers to this as naive modularization, initially built for the scalability of Ethereum as a monolithic Layer 1, with Layer 2 focusing on execution while Layer 1 provides other key functionalities.
Figure 3: Execution, settlement and DA layer architecture
Execution, settlement, and data availability: Optimized modular blockchains can decouple the execution, settlement, and data availability layers across specialized modular blockchains.
Figure 4: Execution and DA layer architecture
Execution and DA: Since the purpose of implementing modular blockchain is flexibility, the execution layer is not limited to simply publishing its blocks to the settlement layer. For example, a modular stack can be created that does not involve the settlement layer, only the execution layer above the consensus and data availability layers.
In this modular stack, the execution layer would be sovereign, publishing its transactions to another blockchain, typically used for ordering and data availability but handling its own settlement. In the context of the modular stack, sovereign Rollup is responsible for execution and settlement, while the DA layer handles consensus and data availability.
The difference between sovereign Rollup and smart contract Rollup is:
Currently, Rollkit and Sovereign SDK provide frameworks for deploying sovereign Rollup testnets on Celestia.
Before introducing modularization of the execution layer, we should understand what Rollup technology is.
Currently, modularization technology of the execution layer relies heavily on Rollup, which is a scaling solution that operates off-chain from Layer 1. This solution executes transactions off-chain, meaning it occupies less block space and is one of Ethereum’s important scaling solutions. After executing transactions, it sends a batch of transaction data or execution proofs to Layer 1 for settlement. Rollup technology provides a scalability solution for Layer 1 networks while maintaining decentralization and security.
Figure 5: Rollup technical architecture
Taking Ethereum as an example, Rollup technology can further enhance performance and privacy by using ZK-Rollup or Optimistic Rollup.
Initially, Ethereum adopted sidechains and sharding technology for scalability, but sidechains sacrificed some decentralization and security to achieve high throughput. The development of Layer 2 Rollups has progressed much faster than expected and has already provided significant scalability, with even more to come after the implementation of Proto-Danksharding. This means that there is no longer a need for “shard chains,” which have been removed from Ethereum’s roadmap.
Ethereum outsources the execution layer to Layer 2s based on Rollup technology to relieve the burden on the main chain, and the EVM provides a standardized and secure execution environment for smart contracts executed on the Rollup layer. Some Rollup solutions are designed with compatibility with the EVM in mind, allowing smart contracts executed on the Rollup layer to still leverage the features and functionalities of the EVM, such as OP Mainnet, Arbitrum One, and Polygon zkEVM.
Figure 6: Ethereum’s layer 2 scaling solution
These Layer 2s execute smart contracts and process transactions but still rely on Ethereum for the following operations:
Settlement: All Rollup transactions are settled on the Ethereum mainnet. Users of Optimistic Rollups must wait for the challenge period to pass or for transactions to be deemed valid after fraud proofs are computed. Users of ZK Rollups must wait for validity to be proven.
Consensus and Data Availability: Rollups publish transaction data to the Ethereum mainnet in the form of CallData, allowing anyone to execute Rollup transactions and reconstruct their state if necessary. Before confirmation on the Ethereum main chain, Optimistic Rollups require significant block space and a 7-day challenge period. ZK Rollups provide instant finality and store data available for verification for 30 days, but require significant computational power to create proofs.
B² Network is the first ZK-Rollup on Bitcoin, enabling increased transaction speeds without sacrificing security. Leveraging Rollup technology, B² Network provides a platform for running Turing-complete smart contracts for off-chain transactions, thereby improving transaction efficiency and minimizing costs.
Figure 7: B² Network architecture
As shown in the diagram, the ZK-Rollup layer of B² Network adopts the zkEVM solution, responsible for executing user transactions within the Layer 2 network and producing relevant proofs.
Unlike other Rollups, the B² Network ZK-Rollup consists of multiple components, including the Account Abstraction Module, RPC Service, Mempool, Sequencers, zkEVM, Aggregators, Synchronizers, and Prover. The Account Abstraction Module implements native account abstraction, allowing users to programmatically incorporate higher security and better user experience into their accounts. zkEVM is compatible with EVM and can also assist developers in migrating DApps from other EVM-compatible chains to B² Network.
Synchronizers ensure that information is synchronized from B² nodes to the Rollup layer, including sequence information, Bitcoin transaction data, and other details. B² nodes act as off-chain validators and executors of several unique functions within the B² Network. The Bitcoin Committer module in B² nodes constructs a data structure to record B² Rollup data and generates a Tapscript called “B² ciphertext.” Then, the Bitcoin Committer sends a UTXO of one satoshi to a Taproot address containing the $B^{2}$ ciphertext, and the Rollup data is written into Bitcoin.
Additionally, the Bitcoin Committer sets a time-locked challenge, allowing challengers to dispute the commitment of zk proofs. If there are no challengers during the time lock or if the challenge fails, the Rollup is eventually confirmed on Bitcoin; if the challenge is successful, the Rollup is rolled back.
Whether it’s Ethereum or Bitcoin, Layer 1 is essentially a single chain that receives extended data from Layer 2. In most cases, the capacity of Layer 2 also depends on the capacity of Layer 1. Therefore, the implementation of the Layer 1 and Layer 2 stack is not ideal for scalability. When Layer 1 reaches its throughput limit, Layer 2 is also affected, which may lead to increased transaction fees and longer confirmation times, affecting the efficiency and user experience of the entire system.
In addition to Celestia’s DA solution being favored by Layer 2s, other innovative solutions focused on DA have emerged, playing crucial roles in the entire blockchain ecosystem.
EigenDA is a secure, high-throughput, and decentralized DA service inspired by Danksharding. Rollup can publish data to EigenDA to achieve lower transaction costs, higher transaction throughput, and secure composability throughout the EigenLayer ecosystem.
When building decentralized temporary data storage for Ethereum Rollup, data storage can be directly handled by EigenDA operators. Operators participate in network operation, responsible for processing, verifying, and storing data, and EigenDA can horizontally scale with the increase in staking and operators.
EigenDA combines Rollup technology while moving the DA part off-chain for scalability. As a result, actual transaction data no longer needs to be replicated and stored on every node, reducing bandwidth and storage demands. On-chain, only metadata related to data availability and accountability mechanisms are processed (accountability ensures data is stored off-chain and can be verified for integrity and authenticity when necessary).
Figure 8: Basic data flow of EigenDA
As shown in the diagram, Rollup writes transaction batches to the DA layer. Unlike systems that use fraud proofs to detect malicious data, EigenDA divides the data into blocks and generates KZG commitments and multi-reveal proofs. EigenDA requires nodes to download only a small amount of data [O(1/n)], rather than downloading the entire blob. The fraud arbitration protocol of Rollup can also verify whether the blob data matches the KZG commitments provided in the EigenDA proof. Through this verification, Layer 2 chains can ensure that transaction data of the Rollup state root is not manipulated by sequencers/proposers.
Nubit is an scalable, Bitcoin-native DA layer aiming to advance the future of Bitcoin by enhancing data throughput and availability services to meet the growing demands of the ecosystem. Their vision is to integrate the vast developer community into the Bitcoin ecosystem, providing them with scalable, secure, and decentralized tools.
The team behind Nubit consists of professors and PhD students from UCSB (University of California, Santa Barbara), with outstanding academic reputations and global influence. They are not only proficient in academic research but also have rich experience in implementing blockchain engineering. The team, along with domo (the creator of Brc20), co-authored a paper on modular indexers, incorporating the design of the DA layer into the indexer structure of the Bitcoin meta protocol, contributing to the establishment and formulation of industry standards.
Nubit’s core innovations lie in its consensus mechanism, trustless bridging, and data availability, utilizing innovative consensus algorithms and the Lightning Network to inherit Bitcoin’s fully censorship-resistant characteristics and improve efficiency through DAS:
Consensus Mechanism: Nubit explores an efficient consensus based on PBFT (Practical Byzantine Fault Tolerance) supported by SNARKs for signature aggregation. The combination of PBFT with zkSNARK technology significantly reduces the communication complexity of verifying signatures between validators, verifying the correctness of transactions without accessing the entire dataset.
DAS: Nubit’s DAS is achieved by multiple rounds of random sampling of small portions of block data. Each successful sampling round increases the probability of complete data availability. Once the predetermined confidence level is reached, the block data is considered accessible.
Trustless Bridge: Nubit utilizes a Trustless Bridge leveraging the Lightning Network‘s payment channels. This approach not only aligns with local Bitcoin payment methods but also does not impose additional trust requirements. Compared to existing bridging solutions, it brings lower risks to users.
Figure 9: Basic components of Nubit
Let’s further review the complete system lifecycle as depicted in Figure 8 using a specific use case. Suppose Alice wants to complete a transaction using Nubit’s DA service (Nubit supports various data types, including but not limited to ciphertext, Rollup data, etc.).
The light client undergoes the DAS process to verify data availability. Additionally, after proposing a threshold number of blocks, checkpoints of this history are recorded on the Bitcoin blockchain through Bitcoin timestamps. This ensures that the validator set can thwart potential remote attacks and support rapid unbinding.
In addition to focusing on chains with specific modular layers, decentralized storage services can provide long-term support for the DA layer. There are also some protocols and chains that offer developers customized and full-stack solutions, allowing users to easily build their own chains without even needing to write code.
EthStorage is the first modular Layer 2 to achieve dynamic decentralized storage, offering DA-driven programmable key-value (KV) storage. It @ld-capital/%E4%BB%8Eethstorage-%E5%9B%9E%E7%9C%8B%E8%A2%AB%E5%B8%82%E5%9C%BA-%E5%86%B7%E8%90%BD-%E7%9A%84%E5%8E%BB%E4%B8%AD%E5%BF%83%E5%8C%96%E5%AD%98%E5%82%A8%E8%B5%9B%E9%81%93-d0a003220362">extends programmable storage to hundreds of TB or even PB at 1/100 to 1/1000 of the cost. EthStorage provides a long-term DA solution for Rollups and opens new possibilities for fully on-chain applications in gaming, social networks, AI, and more.
Figure 10: Application scenarios of EthStorage
Qi Zhou, the founder of EthStorage, has been fully dedicated to the Web3 industry since 2018. He holds a Ph.D. from the Georgia Institute of Technology and has previously worked as an engineer at top companies such as Google and Facebook. His team has also received support from the Ethereum Foundation.
As one of the core features of the Ethereum Cancun upgrade, EIP-4844 (also known as Proto-dank sharding) introduces temporary data blocks (blobs) for Layer 2 Rollup storage, enhancing the network’s scalability and security. The network does not need to verify every transaction in the block, only to confirm whether the blob attached to the block carries the correct data, significantly reducing the cost of Rollups. However, blob data is only temporarily available, meaning it will be discarded within a few weeks. This has a significant impact: Layer 2 cannot unconditionally derive the latest state from Layer 1. If a piece of data can no longer be retrieved from Layer 1, it may be impossible to sync the chain via Rollup.
With EthStorage as a long-term DA storage solution, Layer 2s can access complete data from its DA layer at any time.
Technical features:
EthStorage enables decentralized dynamic storage: Existing decentralized storage solutions can support large data uploads but cannot modify or delete them, only re-upload new data. EthStorage significantly enhances data management flexibility by achieving CRUD functionality (Create, Read, Update, Delete) through an original key-value storage paradigm.
Layer 2 decentralized solutions based on the DA layer: EthStorage is a modular storage layer that can run on any blockchain with EVM and DA to reduce storage costs (though many Layer 1s currently lack a DA layer), and it can even run on Layer 2.
High integration with ETH: The EthStorage client is a superset of the Ethereum client Geth, meaning that when running an EthStorage node, it can still participate in any Ethereum processes. A node can be both an Ethereum validator node and an EthStorage data node simultaneously.
EthStorage’s workflow:
AltLayer offers a versatile, no-code Rollups-as-a-Service (RaaS) solution. Designed for a multi-chain and multi-virtual machine world, the RaaS product supports both EVM and WASM. It also supports various Rollup SDKs, such as OP Stack, Arbitrum Orbit, Polygon zkEVM, ZKSync’s ZKStack, and Starkware, as well as different shared sequencing services (e.g., Espresso and Radius) and different DA layers (e.g., Celestia and EigenLayer), along with many other modular services for different layers of the Rollup stack.
With AltLayer, a multifunctional Rollup stack can be achieved. For instance, a Rollup designed for an application can be built using Arbitrum Orbit, with Arbitrum One as the DA and settlement layer. Meanwhile, a general-purpose Rollup can be constructed using the ZK Stack, with Celestia as the DA layer and Ethereum as the settlement layer.
Note: You might wonder why the settlement layer can be implemented by OP and Arbitrum. In fact, these Layer2 Rollup stacks are currently achieving interchain connectivity similar to what Cosmos proposed. OP has introduced the Superchain, where the OP Stack acts as the standardized development stack supporting Optimism technology, integrating different Layer2 networks and promoting interoperability among them. Arbitrum has proposed the Orbitchain strategy, which allows Layer3s, also known as application chains, to be created and deployed on the Arbitrum mainnet based on the Arbitrum Nitro tech stack. Orbit Chains can settle directly on Layer2s or directly on Ethereum.
Dymension is a modular blockchain network based on the Cosmos SDK, designed to ensure the security and interoperability of RollApps using the IBC standard. Dymension divides blockchain functionalities into multiple layers, with the Dymension Hub serving as the settlement and consensus layer, providing security, interoperability, and liquidity for RollApps, which act as the execution layer. The data availability (DA) layer is supported by Dymension protocol’s DA providers, allowing developers to choose the appropriate DA provider based on their needs.
The settlement layer (Dymension Hub) maintains a registry of RollApps and their critical information, such as state, sequencer lists, currently active sequencers, and execution module checksums. Rollup service logic is fixed within the settlement layer, forming a native interoperability hub. The Dymension Hub, as the settlement layer, has the following features:
In the previous section, we delved into modular blockchain systems and numerous representative projects. Now, we will shift our focus to comparative analysis among different ecosystems, aiming for an objective and comprehensive understanding of modular blockchain.
As we’ve seen, blockchain ecosystems are evolving towards modularity. In the past, blockchain networks operated in isolation, competing with each other, making it difficult for users, developers, and assets to move between different chains, thus restricting the overall development and innovation of the ecosystem. In the Web3 world, identifying and solving problems is a collaborative process. Initially, Bitcoin and Ethereum attracted significant attention as single chains, but as the limitations of single chains became apparent, modular chains began to gain attention. Therefore, the emergence of modular chains is not accidental but rather an inevitable development.
Modular blockchain enhances the flexibility and efficiency of chains by allowing individual components to be independently optimized and customized. However, this architecture also faces challenges such as increased communication latency and complexity in system interactions. In practice, the long-term benefits of modular architecture, such as improved maintainability, reusability, and flexibility, often outweigh its short-term performance losses. In the future, as technology advances, these issues will find better solutions.
GeekCartel believes that blockchain ecosystems have a responsibility to provide reliable foundational layers and common tools throughout the entire modular stack to facilitate seamless connections between chains. If ecosystems can be more harmonious and interconnected, users will be able to more easily use blockchain technology, attracting more new users to Web3.
Currently, there are several Restaking protocols emerging, effectively aggregating dispersed security resources through a re-staking mechanism to enhance the overall security of blockchain networks. This process not only addresses the issue of fragmented security resources but also strengthens the network’s defense against potential attacks while providing additional incentives for participants, encouraging more users to engage in network security maintenance. In this way, Restaking protocols open up new avenues for improving network security and efficiency, powerfully promoting the healthy development of blockchain ecosystems.
EigenLayer is a protocol built on Ethereum that introduces the Restaking mechanism, which is a new primitive for cryptographic economic security. This primitive allows the reuse of ETH at the consensus layer, aggregating ETH security across all modules, thus enhancing the security of DApps that rely on these modules. Users who stake native ETH or use Liquidity Staking Tokens (LST) to stake ETH can choose to join the EigenLayer smart contract to restake their ETH or LST, extending cryptographic economic security to other applications on the network and earning additional rewards.
As Ethereum moves towards a Rollup-centric roadmap, applications built on Ethereum are experiencing significant scalability. However, any module that cannot be deployed or proven on the Ethereum Virtual Machine (EVM) cannot absorb the collective trust of Ethereum. Such modules involve processing inputs from outside Ethereum, making their processing unverifiable within Ethereum’s internal protocols. These modules include sidechains based on new consensus protocols, data availability layers, new virtual machines, oracle networks, bridges, and more. Typically, such modules require an @GenesisLRT/what-are-avss-and-operators-00e1c51dab1c">Autonomous Verification System (AVS) with its own distributed verification semantics for validation. Usually, these AVSs are protected by their native tokens or have permissioned properties.
Currently, there are some issues within the AVS ecosystem:
Figure 11: Comparison between current AVS service and EigenLayer
In the architecture of EigenLayer, AVS (Autonomous Verification System) is a service built on the EigenLayer protocol, leveraging the shared security of Ethereum. EigenLayer introduces two novel approaches, through staking and free market governance, to achieve centralized security. These approaches help extend Ethereum’s security to any system and eliminate the inefficiency of existing rigid governance structures:
By combining these approaches, EigenLayer acts as an open market where AVSs can leverage the pooled security provided by Ethereum validators, encouraging validators to make more optimized trade-offs between security and performance through reward incentives and penalties.
Babylon is a Layer1 blockchain founded by Stanford University professor David Tse. The team consists of researchers from Stanford University, experienced developers, and business advisors. Babylon introduces the Bitcoin staking protocol, designed as a modular plugin to be used on various PoS (Proof of Stake) consensus algorithms, providing a primitive for re-staking.
Babylon leverages three aspects of Bitcoin—timestamping service, block space, and asset value—to propagate Bitcoin’s security to many PoS chains (such as Cosmos, Binance Smart Chain, Polkadot, Polygon, and other blockchains with robust and interoperable ecosystems), creating a stronger and more unified ecosystem.
Bitcoin timestamping resolves PoS long-range attacks:
Long-range attacks involve validators in a PoS chain unstaking and going back to a historical block where they were still validators, potentially starting a fork. This problem is inherent in PoS systems and cannot be completely solved solely by improving the consensus mechanism of PoS chains themselves. Both Ethereum and Cosmos, among other PoS chains, face this challenge.
After introducing Bitcoin timestamping, on-chain data of PoS chains will be stored on the Bitcoin chain with Bitcoin timestamps. Even if someone tries to create a fork of a PoS chain, the corresponding Bitcoin timestamp will definitely be later than the original chain, rendering the long-range attack ineffective.
Bitcoin staking protocol:
This protocol allows Bitcoin holders to stake their idle Bitcoin to enhance the security of PoS chains and earn rewards in the process.
The core infrastructure of the Bitcoin staking protocol is the Control Plane between Bitcoin and PoS chains, as illustrated in the diagram below.
Figure 12: System architecture with Control plane and Data plane
The Control Plane is implemented in the form of a chain to ensure it is decentralized, secure, censorship-resistant, and scalable. This control plane is responsible for various critical functions, including:
• Providing Bitcoin timestamping services for PoS chains to synchronize with the Bitcoin network.
• Acting as a marketplace, matching Bitcoin staking with PoS chains and tracking staking and validation information, such as the registration and refresh of EOTS (Epoch Time Oracle Service) keys.
• Recording the finality signatures of PoS chains.
By staking their BTC, users can provide validation services for PoS chains, DA layers, oracles, AVSs (Autonomous Verification Systems), etc. Additionally, Babylon can now offer services for Altlayer, Nubit, and other platforms.
Figures:
Articles:
In this emerging infrastructure paradigm, there is still much research and work to be done, and there are many areas that this article did not cover. If you are interested in any related research topics, please contact Chloe.
A big thanks to Severus and Jiayi for their insightful comments and feedback on this article.
This article is reproduced from [medium], the original title is “Modular Blockchain: The Last Piece of the Web3 Puzzle”, the copyright belongs to the original author [GeekCartel], if you have any objection to the reprint, please contact Gate Learn Team, the team will handle it as soon as possible according to relevant procedures.
Disclaimer: The views and opinions expressed in this article represent only the author’s personal views and do not constitute any investment advice.
Other language versions of the article are translated by the Gate Learn team, not mentioned in Gate.io, the translated article may not be reproduced, distributed or plagiarized.