Modular Architecture and its Impact on the Development of the Web3 Industry: How Should Ethereum Respond?
Blockchain technology has witnessed the emergence of two different design philosophies, namely monolithic architecture and modular architecture, each with its own advantages and application scenarios. Solana’s monolithic architecture offers fast and cost-effective transaction experiences, but its high hardware requirements may lead to centralization trends. On the other hand, modular designs such as Celestia and Ethereum’s latest developments reduce costs, enhance security and flexibility through the separation of data availability layers, and support a wider range of applications and innovations.
These developments reflect the continuous evolution of blockchain technology and the ongoing exploration of the balance between security, decentralization, and scalability by the community. With further advancements in modular architecture, we may witness a more diverse and vibrant blockchain ecosystem, providing users and developers with more choices and possibilities. Despite different perspectives and approaches, these advancements collectively drive the progress of blockchain technology, marking its evolution towards a more mature and flexible framework.
Recently, Solana has gained significant popularity for good reasons. It has emerged from the dark days of the Alameda crisis and has shown strong momentum, successfully handling the busiest airdrop claims in history while maintaining extremely low fees. From the perspective of attracting new users, Solana is a good choice, as Ethereum’s layer 2 still charges transaction fees as high as $1 (starting from BSC or Tron may not be a good idea, in our opinion).
Another advantage of Solana is its unified global state, which allows real-time response to all market signals without arbitrage and bridging between rollups or shards. It’s like trading seamlessly across all global exchanges, with events reflecting instantaneously in the price changes across all exchanges, regardless of geography or time zone. These are the benefits of a single-chain best-case scenario. However, this design choice still has its drawbacks, notably the tendency towards centralization due to its high hardware requirements. Solana validators tend to centralize as it handles all three layers of the blockchain (execution, consensus, and data availability) in a single manner.
On the other end of the design spectrum, modular architecture, especially outsourcing the data availability layer, is gaining popularity. This approach reduces transaction costs while maintaining low hardware requirements (although MEV poses a threat to this). Modular designs also allow for specialized chains and hardware for specific applications, with dYdX being the best example.
At the forefront of the modular movement is Celestia, a chain optimized for rollup data efficiency. On the other hand, Ethereum has achieved modularity in a more incremental way, building the plane while flying. We believe that rollups are the key to scalability and cost reduction, and now the battle for the data availability layer (and the rest of the modular stack) has begun.
The data availability problem was initially identified in the early scalability race of blockchain. The focus at the time was on minimizing the amount of stored data to maximize the number of nodes in the network. This dynamic also formed the basis for the Bitcoin block size war. Data availability refers to the ability of a blockchain to make its data accessible to all network participants. The key breakthrough in solving this problem was the introduction of Data Availability Sampling (DAS), as explained by Bridget Harris:
“Through DAS, light nodes can confirm data availability by participating in randomly sampled rounds of block data, without having to download every complete block. Once multiple rounds of sampling are completed and a certain confidence threshold is reached, data availability is confirmed, and the rest of the transaction process can proceed securely. This way, a chain can expand its block size while maintaining verifiable data availability. It also achieves significant cost savings: these emerging layers can reduce DA costs by up to 99%.”
Celestia, Avail, NearDA, and EigenDA are the most notable DA projects. They do not require the validation of transactions but only check if each block has been added through consensus and if the new block is available to the network. They rely on third-party sequencers to execute and validate transactions. Celestia launched in October 2023, Avail and EigenDA will launch their mainnets in the coming months, and Near recently announced its DA solution. Let’s review the unique features of each project:
1. Celestia chose the fastest path to market, adopting fraud proofs (also used by Optimism rollups). The tradeoff is that, under the current configuration, Celestia will not be able to support ZK rollups. The Celestia team claims that approximately 70% of all new Arbitrum Orbit chains are using Celestia for data availability.
2. Avail (formerly Polygon Avail) serves as an independent blockchain that provides a fast and secure data and consensus layer, giving developers what they need to launch rollups, whether ZK or optimistic.
3. EigenDA is possibly the most Ethereum-compatible as it is a DA module rather than a blockchain. Additionally, ETH staked in the EigenLayer will be available to protect rollups using EigenDA. Its weakness is that it does not use data sampling or data availability proofs.
4. NearDA helps rollups save data availability costs by storing data on the Near sharded blockchain. NearDA leverages an important part of the Near consensus mechanism, which parallelizes the network into multiple shards.
Now let’s talk about rollups themselves. Within the rollups built on top of these data availability (DA) providers, there are many tools to make rollup launch easier:
1. Manta Pacific offers significantly reduced costs compared to monolithic L2 solutions by leveraging modular data availability from Celestia, saving $1 million in Ethereum gas fees. Manta also uses custom opcodes to verify ZK technology, making privacy protection and native randomness significantly cheaper in their protocol.
2. Mantle Network, built on a modular architecture, combines optimistic rollup protocols with EigenDA’s data availability solution. This integration allows the Mantle Network to inherit Ethereum’s security while providing more economical and accessible data availability.
3. Kinto is a KYC-required chain where every user and developer on the network must complete a passport KYC process before transacting. It uses Celestia to lower costs.
In a truly modular fashion, modules from each layer are chosen based on specific needs. The diversity of possible combinations can be seen here:
Projects like Eclipse, a “Rollup-as-a-Service,” make rollup launches easier, allowing developers to choose which technologies to use for each module.
Similarly, Conduit enables users to deploy a rollup within 15 minutes, supporting Optimism, Arbitrum Orbit, and Celestia. Users pay monthly fees for the hosting infrastructure to Conduit and separately for data availability to the provider.
The richness of possible combinations created by modularity is undoubtedly a significant advancement. Is this akin to the difference between the difficulty of building a website in the past compared to the convenience and customization offered by platforms like Squarespace today?
This statement means that modular technology simplifies the complexity of technical implementation by providing various combinations, similar to the convenience and customization offered by platforms like Squarespace compared to the difficulty of building a website in the past. Squarespace and similar modern platforms allow people without technical backgrounds to easily build and manage websites, and modular technology offers similar convenience and flexibility in the implementation of blockchain and rollups.
While there is skepticism about outsourcing data availability (DA), Ethereum places great importance on modular architecture. The vision initially pursued through sharding for scalability has been abandoned in favor of modularity.
The three key updates needed to realize the vision are rollups (discussed earlier), proposer-builder separation (“block proposers no longer generate blocks for their own “profit-maximizing” purposes but delegate the task to external participants (builders) in the market”), and data sampling. The latter is a way for light nodes to verify block availability by downloading a randomly selected small amount of data. Technically, this is more challenging than the other two and will take two to three years to implement.
It is important to note that EIP-4844 is the first step in improving Ethereum’s data availability layer before data sampling goes live. As discussed earlier, improving Ethereum is like repairing an airplane while flying. Once the Ethereum Foundation recognized the need for rollups (when Vitalik presented the famous rollup-centric future), the team chose to expand the block package by adding blobs, a space specifically tailored for rollup data. Blobs are expected to reduce rollup transaction costs by a factor of ten. EIP-4844 is planned to go live in the Dencun upgrade in March/April. While it is a temporary solution to keep Ethereum competitive within the next two to three years, the long-term solution will be supporting validity proofs on the mainnet itself, which will reduce the cost of rollups by several orders of magnitude.
Although Solana may staunchly defend its monolithic architectural philosophy (which may be correct for many use cases), the industry seems to be moving towards modularity. For Ethereum, only a modular architecture can achieve the future:
1. Cost-effective transactions for millions of users through rollups (scalability).
2. Protection against threats like censorship and 51% attacks (security).
3. Ability to run a node on an ordinary PC or even a smartphone to validate transactions (decentralization).
One might ask if Ethereum’s modular architecture solves the so-called blockchain trilemma that was once considered unsolvable. Technically, it does not, as Ethereum is no longer a monolithic network. However, as a modular network, it does achieve this.
Among these three, we believe that solving decentralization is the most important part of the trilemma. Innovation will eventually reduce transaction costs, and prioritizing decentralization, especially geographical distribution, is the only way to ensure long-term network security. Ethereum leads in decentralization, with the most decentralized set of validators, exceeding 800,000 validators. At the same time, through a modular approach, it can adapt to new design innovations by launching custom rollups on top. Celestia and others undoubtedly share this vision too. The remaining question is whether Ethereum can move fast enough in this modular direction to keep up with competitors building from scratch rather than repairing the plane while flying.
Related Reports
Monolithic Chains vs. Modular Blockchains: The Next Steps for Ethereum, Solana, and Celestia in the Competitive Landscape
In-depth Explanation of Celestia’s Core and Economic Model: Explosive Growth of Modular Blockchains?
Celestia is Live! Token TIA Surpasses $2, Two-Step “Update Keplr Wallet” for Airdrop Claim