TL;DR

AI dominates as the trendiest sector in the venture capital scene, closely followed by Bitcoin-centric initiatives. On any given day, discussions about these topics can dominate up to 80% of all project talks, with personal records reaching five to six AI projects.

The AI sector is expected to hit its speculative peak within the next couple of years. Despite the eventual bursting of this bubble, which will likely cause significant market disruptions, it’s predicted that this phase will also birth unicorns that successfully marry AI with cryptocurrency, driving substantial forward momentum in the industry.

In light of the current buzz around AI, it’s beneficial to pause and reflect on the infrastructural evolutions within public blockchains over the past months. Some innovations in this area are particularly noteworthy and merit closer examination.

1. The Evolution and Expansion of Ethereum and Single-Layer Blockchains

The concepts of modularity and Data Availability (DA) layers, first proposed by Celestia, have now become a cornerstone of blockchain infrastructure, deeply ingrained in the community’s consciousness. This has led to an explosion in the number of Rollup-as-a-Service (RaaS) infrastructures, now outnumbering both the applications they support and their users—a stark indication of the market’s saturation.

In recent months, there have been notable technological advancements across the execution, DA, and settlement layers of blockchain infrastructure, each layer evolving with its own set of new solutions. The settlement layer, once dominated by Ethereum, is now more competitively diverse.

1. Developments in the Execution Layer

The concept of Parallel EVM, spearheaded by projects like Monad, Sei, and MegaETH, stands out as the most groundbreaking in the execution layer. Existing projects such as FTM and Canto are adapting to embrace this innovative direction. However, it’s important to note that not all projects associated with the Parallel EVM share the same technological paths or end goals.

For instance, Sei’s diagrams vividly illustrate the potential performance boosts achievable by moving from sequential to parallel processing under ideal conditions.

Parallel EVM technologies can be categorized based on their transaction parallelization approach:

Pre-validation methods, such as those used by Solana and Sui, require transactions to specify which parts of the chain state they modify, allowing for pre-block packaging conflict detection and the discarding of conflicting transactions.

Post-validation methods, known as optimistic parallelism and exemplified by Aptos’ BlockSTM, assume no initial conflicts, processing transactions first and resolving any detected conflicts afterwards by invalidating conflicting transactions and reprocessing as necessary. This method is also employed by Sei, Monad, MegaETH, and Canto.

There are also emerging solutions designed to handle state conflicts, such as those involving simultaneous access to the same Automated Market Maker (AMM) pool, although these solutions appear more complex and their commercial viability remains under assessment.

Perspectives on Parallel EVM -

There are two main perspectives on the importance of Parallel EVM:

The first perspective, championed by projects like Monad and Sei, places transaction parallelization at the forefront of their scalability strategy. Monad, for instance, not only advocates for optimistic parallel processing but has also developed specialized tools like MonadDB and asynchronous I/O to support these efforts.

The second perspective, represented by Fantom, Solana, and MegaETH, views parallelization as one of several scalability strategies, not the sole focus. These projects also rely on other technological advancements to enhance performance.

For example, Fantom’s Sonic upgrade is centred around its FVM virtual machine and an improved Lachesis consensus mechanism. Solana’s upcoming initiatives focus on the modular architecture of its Firedancer client and enhancements in network communications and signature validations.

MegaETH @megaeth_labs aims to push the boundaries of Ethereum’s capabilities towards achieving a near “Realtime Blockchain,” enhancing various aspects like state synchronization, hardware configurations for Sequencers, and the data structure of the Merkle Trie, all aimed at maximizing the efficiency and speed of blockchain operations.

1. Data Availability (DA) Layer

The DA layer hasn’t seen any major technological iterations, so the degree of development here is not as intense as in the execution layer. Essentially, there are only a few key players:

Ethereum’s CallData upgrade to Blob has significantly reduced costs for various Layer 2s, making ETH a “not so expensive” option for DA now.

Celestia’s major role, after its launch, was as the first project to introduce the DA layer concept. This project increased the ceiling for the DA track from $2 billion in Fully Diluted Valuation (FDV) to $20 billion, opening up new frameworks and imaginative possibilities. Many new Layer 2 Appchains naturally prefer Celestia for their DA.

Avail, which spun off from Polygon, technically resembles an “enhanced version of Celestia,” such as using the Grandpa+BABE consensus mechanism similar to Polkadot’s, theoretically supporting more decentralized nodes than Celestia’s Tendermint. It also supports Validity Proofs which Celestia does not. Of course, technical differences are not as critical as ecosystem development, and Avail needs to catch up in terms of its ecosystem.

EigenDA was also launched along with the EigenLayer mainnet a few days ago. With EigenLayer being one of the strongest narratives this round and best at business partnerships, I feel that the adoption rate for EigenDA will be high. Theoretically, as long as it feels secure and the price is right, not many projects care whether you use Validity Proof or Fraud Proof, or whether DAS is supported, etc.

It’s worth mentioning the following three DAs:

• Near DA - Near is a unique blockchain originally focused on sharding, and it still does, but while doing so, it also developed a DA - cheaper than Celestia and supporting fast settlement for L2s. Chain abstraction - Recently, Near introduced chain signatures, allowing users to request transaction signatures for any blockchain via a single NEAR account; AI - Founder Illia, one of the “eight sons of Transformer” and who was patted on the shoulder by Jensen Huang at the NVIDIA conference, is now planning to hire AI engineers and will publish announcements related to http://near.ai next month… I’ve also placed it in the DA track.
• BTC & CKB - Because BTC’s first layer doesn’t support smart contracts and can’t settle directly, dozens of BTC EVM Layer 2s are essentially using BTC as DA. The difference is whether to put the ZK Proof directly on BTC or just the hash of the ZK Proof, as if not doing so would disqualify it as a “BTC Layer 2.” I recently came across a new project saying “I’m not pretending anymore, I’m an ETH L2, settlements are done on ETH, but I serve the BTC ecosystem!”, which is quite humorous… The only alternative expansion scheme is CKB’s RGB++, where CKB has become a kind of DA existence, and BTC, due to UTXO homomorphic binding technology, has almost become the settlement layer for RGB++.
• New DA - Discuss two new DA ideas I’ve seen, not naming the projects: one combines DA with AI, not only serving as a high-performance DA but also acting as a storage layer for AI large model training data and trajectories. Another improves on the error-correction code mechanism underlying DAs like Celestia, providing more robust network stability in dynamic networks (where a few nodes randomly go offline each round).

1. Settlement Layer

Originally, this layer was almost exclusively dominated by ETH, with DA having competition from Celestia, and execution having its plethora of L2s. Only in settlement, other chains like Solana, Aptos, etc., do not have L2s, and BTC’s L2s can’t use BTC for settlement, so the only settlement layer you can think of is essentially ETH alone.

However, this situation is about to change. I’ve seen several new projects moving in the direction mentioned at the beginning of the article, and some old projects are also pivoting in this direction, i.e., ZK verification/settlement layer - further deconstructing ETH (stealing ETH’s business).

Why has this concept emerged?

From a technical perspective, running contracts on ETH L1 to verify ZK Proofs is indeed not the optimal choice. To verify the correctness of ZK Proofs, developers need to write verification contracts in Solidity based on the ZK project and its chosen ZK Proof System, which involves complex cryptographic algorithms such as different elliptic curves. These algorithms are generally complex, and the EVM-Solidity architecture is not the optimal platform for implementing these complex cryptographic algorithms. For some ZK projects, writing and verifying these verification contracts is also costly.

This somewhat hinders the native integration of some ZK ecosystems into the EVM ecosystem, hence languages like Cairo, Noir, Leo, and Lurk can currently only be verified on their Layer 1s. Also, updating or upgrading such things on ETH is always “hard to turn around a big ship.”

From a cost perspective, although the “protection fees” on L2 account for the bulk, ZK contract verification also requires Gas fees, and Ethereum is certainly not a cheap option for verification. With ETH Gas fees occasionally soaring, turning it into a “noble chain,” the verification cost is also greatly affected.

Thus, new ZK verification/settlement layer concept projects have appeared, still relatively early, with Nebra as a representative. Old projects are also pivoting in this direction, such as Mina and the recently passed new proposal by Zen.

Most projects in this track generally aim to:

• Support multiple ZK languages.
• Support aggregated ZK proofs, which are more efficient and cheaper.
• Achieve faster finality times.

It’s quite likely that the ZK settlement layer and the decentralized Proof Market will be linked, as having the technology also requires computational power. We may see some settlement layer projects cooperate with Proof Market projects, powerful settlement layers might start their own Proof Markets, or technically proficient Proof Markets might enter the settlement layer arena themselves. Ultimately, the market will decide.

Other areas of Infra, such as the Oracle and MEV fields with OEV, and the interoperability field with ZK light clients, are well covered in online articles, which I will not elaborate on here. Next time I see some new and interesting things, I’ll share them with everyone.

Disclaimer：

1. This article is reprinted from [Twitter], All copyrights belong to the original author [@Wuhuoqiu]. If there are objections to this reprint, please contact the Gate Learn team, and they will handle it promptly.
2. Liability Disclaimer: The views and opinions expressed in this article are solely those of the author and do not constitute any investment advice.
3. Translations of the article into other languages are done by the Gate Learn team. Unless mentioned, copying, distributing, or plagiarizing the translated articles is prohibited.