From bridge liquidity to chain abstraction full-stack framework, what innovations have we seen?

Chain Abstraction

Humans are not usually asset managers, but they are forced to become money managers. The average number of bank accounts per person in the US is between 3 and 4, suggesting that even in a highly developed banking system, the average American only manages about 4 accounts, despite the seamless movement of money between them.

15 of the top 25 projects by market cap are “consumer-focused L1” projects. Will the average non-crypto consumer dabble in more than 3-4 chains? Do they need to treat their gas tokens as assets, too?

With the rapid growth of the crypto market, I think the market is ready to choose their favorite three chains (including rollups). Why would they need to do that if the technology is advancing?

Chain abstraction is the end game for liquidity fragmentation in crypto. DEX aggregators win when there is liquidity fragmentation across different DEXs; bridge aggregators emerge when there is liquidity fragmentation across different bridges for interchain operations (of course, bridge aggregators contribute much more to the ecosystem than that); and finally, when liquidity is fragmented across different chains, the intuitive sense of chain abstraction becomes very clear. We said from the beginning that "one day, users won't even know which chain they are using."

This is why I am bullish on the concept of chain abstraction. It will help to make crypto participation much more accessible without the mental burden of maintaining multiple accounts on multiple chains. This article will dive into the implementation of chain abstraction, the pros and cons, the trade-offs, and the likely winners in the end.

I define chain abstraction as: any intent initiated by a user on the chain of their choice (where the liquidity is) and executed on the chain where the application is located (where the result is).

The user submits an intent on chain A, and after some magical operation, the user gets the desired result on the target chain or back in the same wallet.

This "magic" can be achieved in a number of different ways, involving different trust assumptions, adoption curves (both user and developer), and goals for the chain abstraction experience that the end application hopes to unlock. Different projects have different views on chain abstraction, but here are the key layers to achieve this goal. Different projects are addressing different layers, and after reading this article, you should have a fair understanding of the key elements required and what the final state may look like.

Interaction layer

These projects try to abstract the chain from the user's first interaction.

To the user, this may look like a multi-chain version of the account abstraction, in the form of a wallet or a

unified front-end interface for interacting with multiple chains (such as a cross-chain lending platform).

Projects built in this direction include: NEAR, Particle Network, Light.

NEAR Protocol

NEAR Protocol aims to simplify the abstraction of blockchains for users as much as possible. They have relayers to subsidize gas fees, an authentication service to recover accounts via email (very similar to Web2's user experience), and most importantly, multiple signature types generated using NEAR accounts.

Applications can stay where they are, with almost no development costs other than integrating the NEAR wallet.

Multiple signature types help users interact with multiple chains at the same time. While this sounds simple, solving liquidity and messaging is very important. NEAR must be able to connect to multiple chains through single or multiple messaging protocols and liquidity networks.

Since they are closest to users, NEAR must also market aggressively and occupy a high market share.

Particle Network

Particle Network holds a similar view on chain abstraction. They originally started as an AA wallet in the EVM ecosystem, but are now moving towards chain abstraction by creating "universal accounts" on their modular L1. This modular L1 is built using the Cosmos SDK. This enables Particle Network to be IBC compatible with any inter-chain communication. They are also using Berachain's Polaris framework to become EVM compatible with the Cosmos chain.

Particle Network will not rely on any external protocol to provide liquidity. Since they are their own chain, they will optimistically perform cross-chain atomic transactions and have their own Gas token.

We do see a lot of overlap in the approaches of NEAR and Particle. While Particle controls most of the tech stack, they will have the additional task of bootstrapping and maintaining their liquidity network, in addition to similar issues faced by NEAR. Light Light.so is a relatively new project that takes an account abstraction approach, but is limited to the EVM ecosystem. By taking advantage of the typical gas fee abstraction and batch execution, they have transformed the wallet user experience to provide a full dashboard experience. Light is committed to abstracting many common operations and providing users with a dashboard-like experience. Future development paths may include integrating multiple DeFi operations into the dashboard, such as swaps, borrow/lend, structured yield products, etc. However, in order to facilitate these operations, a bridge/messaging layer is still required on the backend.

Communication layer

The interaction layer needs to go through the task execution layer, which can be a bridge, proxy, validator, or any infrastructure that can achieve cross-chain interoperability.

Standardized validator network

Across has taken the lead in cross-chain aggregators. Existing crypto users who frequently use the Ethereum ecosystem may be familiar with Across. Across has transformed into an intent-driven structure in its V2 version, positioning it as a leader in the bridge aggregator war. This also helped enable V3, where developers can conveniently combine bridge and protocol operations in a single transaction.

Hypothetical example: OpenSea integrates Across+. If I want to buy Base Chads on Base, I just need to sign a transaction on Arbitrum using my wallet of choice, and then I can successfully buy a Base Chad from the same address on Base.

This example was the easiest for us to understand because it looked like the solution we were looking for.

This approach is great for quickly buying a selected memecoin or buying an NFT listed on a marketplace, but may not be suitable for high-frequency activities like a Telegram Bot, or signing every action as a transaction on a rollup hosting a poker game. In the latter case, bridging and using a rollup may be easier to achieve lower latency and better execution.

Anoma takes a unique intent-driven off-chain approach, with a validator-based L1 and consensus mechanism. Developers can build directly on Anoma or use Anoma as a middleware (essentially a validator network). To standardize communication within the network, Anoma has its own DSL that developers need to learn in order to leverage Anoma's network.

Validator network standardization is one of the hottest research areas in chain abstraction. Issues such as validator centralization, auction mechanisms, the impact of open validator networks have been debated for a long time, and I will not delve into these issues here. Here is a best article on intent-based bridging architecture by Arjun Chand, which includes risks and trade-offs.

Projects such as Ethereum Swap, UniswapX, and 1inch Fusion have demonstrated first-class execution of intent-based architecture. There is no doubt that intent-based architecture will dominate the cross-chain and chain abstraction space, but who will win? We have seen that order flow is king. The validator network that can guarantee best execution will get the best order flow, no matter where the orders come from. Can chain abstraction wallets give them the best order flow?

How good are validators for high-frequency activity? How good are they for latency-critical transactions (e.g., buying low-liquidity memecoins)? These are probably not the best use cases for validator networks or chain abstraction in general.

The top-level activity that a mature validator network can enable in the chain abstraction paradigm is large-scale cross-chain (e.g., moving ETH from all L2s to a single Ethereum mainnet account). Anywhere there is a need for research overhead, integration overhead, bridging overhead (including aggregators), gas maintenance, etc. is where validator infrastructure can help. Buying Injective derivatives on Injective should be seamless and one-click, and I can do it even if I don't have any funds.

Competitive Landscape of Validator Networks

To ensure execution, each validator network needs to integrate with some contracts. Across V3 is leading with its intent-driven architecture and now just needs to sort out the integration issues with protocols. Protocols will likely integrate with battle-tested projects like Across, and they will need to continue to innovate their architecture to attract more validators (or relayers as they call them) to participate without affecting execution.

However, Across V3 is not a clear winner in terms of order flow. The Stargate bridge is competing head-on with Across in terms of order flow and trading volume, and Celer Circle and cBridge also seem to be catching up.

Across is the only project with an intent-driven architecture that has consistently delivered superior execution. There has long been a belief that Stargate’s volume was artificially inflated through incentives, but there is no way to prove this. However, while Stargate’s volume is comparable to Across’, it has double the number of transactions. Only after the

LayerZero airdrop is complete will we be able to determine which volume is incentivized and which is non-incentivized.

Socket takes a unique approach by introducing a modular order flow auction architecture (MOFA), where any of the above modules can submit an order or participate in an auction. I am not familiar with the underlying technology, but with the team’s track record of shipping great products, this could be very interesting.

Image Courtesy: Socket

Bridges and Bridge Aggregators

“Cross-chain bridges are cumbersome to use” - User Voice

Bridge aggregators used to be my favorite way to transfer assets across chains. It can ensure that assets are bridged to the chain of the user's choice in the best way. Although it is currently the best form of cross-chain transfer, it only blocks the bridge itself and does not block the blockchain. Users still need to hold a minimum amount of gas on the target chain to complete cross-chain transfers. They also don’t help users perform operations on the target chain, which may bring additional complexity to the user experience for users who are new to this space.

At scale, bridges are not as efficient as validator networks. Why? I recommend watching Hart Lambur’s talk at EthDenver 2024 to understand why batching intent can be over 50 times cheaper than traditional bridges. (See 9:11 - 13:25).

While I appreciate the teams and founders who are working to build bridges and allow me to interact in a multi-chain world, I would rather completely eliminate 3-4 steps in the user flow and the slight anxiety that comes with it.

Full-stack frameworks

Full-stack frameworks help create standards from the wallet layer to the settlement layer, and seem to enable complete chain abstraction for users in terms of technical efficiency (security, communication, etc.). Frameworks such as CAKE make protocols easy to adopt and integrate into the entire ecosystem.

It will be very difficult for developers to build projects completely dependent on a brand new framework or chain. The driving force behind developers choosing a specific framework is usually order flow.

I don't know how to convince an entire ecosystem of developers who have already chosen their favorite environment to build their projects to use a brand new framework. This will be a battle that relies heavily on marketing and partnerships, and is as difficult as launching a new L1.

Full stack framework participants include: CAKE, DappOS, Aarc.

Summary

A unified framework is critical, and the leader of each module will be determined by the best order flow. The best order flow depends on continuously providing the best execution results. The entire chain abstraction framework might look like this:

If I had to expose my grandmother to cryptocurrency, I would probably wait until NEAR or Particle Network release a product before taking action. Because I don’t want her to get stuck in the cycle of learning bridges/aggregators, verification, and maintaining multiple private keys, when all she needs is an EVM wallet and buying some token on Solana.

In order to achieve all of this, some form of account abstraction, balance abstraction, and possibly even gas abstraction will be required, and many participants are working on solving each problem.

Based on the information currently available, the leaders of each module will decide the ideal framework. NEAR seems most likely to be the on-ramp for new order flow, Across seems to be the battle-tested project and the easiest to integrate (depending on the Chaos Labs team to further optimize the protocol, they know how to win in a crowded ecosystem), and finally the cross-chain messaging layer, which will provide a secure environment for auxiliary infrastructure (such as bridges and oracles) to provide settlement services for assets moving across chains.