In our July ’21 article ‘What Does the Ideal Cross-Chain Bridge Look Like’, we focused on the trade-off among different asset cross-chain bridges. Here, we made two points regarding the future of cross-chain bridges:
(1) The realization of heterogeneous chain information on the cross-chain basis is extremely difficult.
(2) Current asset cross-chain bridging has already met the needs of DeFi in the short-term.
Obviously, the cross-chain bridges enabling the transfer of token assets by themselves are not enough, concerning the diverse scenarios of cross-channel communication. Furthermore, one could not expect there is a one-size-fits-all solution for the cross-chain bridges for the transfer of token assets.
Multiple blockchain ecosystems are emerging and provide various features appealing to users and application developers. However, few communications happened across different ecosystems. In that case, smooth cross-chain communication is much more important than ever. And cross-chain communication protocol track is likely to usher in an explosion this year.
TL; DR
- The ideal information cross-chain should be as close as possible to the most basic principles of the blockchain.
- All cross-chain bridges all have trade-off among latency, security, complexity, overhead, and reusability, because of the principles of blockchain.
- With the appearance of LayerZero and Axelar, we see the trade-off narrowed through two different innovation paths, which is worth looking forward to.
- Axelar has the potential to become the hub of eco-systems. There will be huge opportunities.
What is information cross-chain?
What is information cross-chain? Let’s gain a basic understanding of it from James Prestwich’s speeches.
Cross-chain, the purpose is that the message sent from Dapp on the original chain could be received from Dapp on destination chain.
Unfortunately, there are varies between chains, we cannot just simply send information from one chain to another. The process cannot happen automatically.
A few more steps are needed:
First, we need to have the receiving chain process the transaction that came from the sending chain. Second, when we are sending info from one chain to another, the only way is to make a transaction. So the sending chain needs to dispatch the transaction.
Unfortunately, we cannot do either, because the blockchain doesn't have the private key. So blockchain cannot sign a transaction from another blockchain. So you cannot send transactions from one chain to another.
What happens in practice is that:
Someone running a server somewhere, sees that message, decides to forward it to the other chain, and signs and dispatches the transaction. It is impossible for one chain to just dispatch a transaction to another chain.
Chains are passive. They cannot reach out outside themselves. It can only update its own state. This is how the blockchain technology works essentially.
An operation in blockchain is a state transition function in basics, simply APPLY(S,TX) -> S' no matter the scenarios are asset transfer, deFi or even the metaverse.
Cross-chain communications, on the other hand, need to be active. To send a message and ensure it would get there, a communication channel is needed. Such a third-party could enable the state to be updated cross-chain.
The ideal cross-chain communication should rely on the most basic elements of blockchain, the event and the state. Upper layer handing such as transaction tracing is prone to errors.
Current Cross-Chain Interoperability
Today, there are nearly a hundred cross-chain bridge projects. Among those, there are three main types of cross-chain interoperability, each with various trade-offs in terms of latency, security, complexity, overhead, and reusability between blockchains with heterogeneous properties, and preserving the atomicity of transactions:
One of the most popular models for bridging used by Matic, and basically every other chains right now is a trusted bridge. Money is handed to trusted custodian, and the custodian can make each message to be received by the remote chain, so the custodian could run off with all the money. However, we need to be alert that the custodian might make smart contracts to receive messages that were never even sent as well.
Interoperability Hubs need years to build. One example, IBC, takes the headers from another chain and verifies them within the local chain. The way a header is verified is specific to the Tindermint BFT consensus PoS protocol, but the header verification can't be reused for other chains with different headers and different consensus protocols. Therefore, in order to connect to IBC, the entire header chain validation logic would have to be rewritten. That makes bridging cosmos zones to outside chains a different story.
Rollup Bridges are designed specifically for connecting mainnet to the rollup. They're usually orchestrated by a contract on the mainnet. They are completely bespoke bridges with relatively low overhead compared to a header relay like IBC but very high latency.
With all that said, none of the three types are sufficiently satisfying. We are actively looking for one that could achieve cross-chain interoperability at a more fundamental level as well as maintain a better equilibrium with an acceptable trade-off.
Among the current cross-chain bridges, the prospect of Axelar and LayerZero is most promising. Those two are believed to be able to stand out among all the other cross-chain bridges. Hence, I'd like to dig deeper into their scenarios, limitations, and visions in the following parts.
LayerZero
LayerZero is a ‘trustless’ omnichain interoperability protocol. It is a lightweight communication layer solution. The developers can use LayerZero protocol as a plug and play developing package and interact through API directly with the Endpoint of LayerZero on its located chain.
Basic Principles:
The interface on LayerZero is a lightweight on-chain client, which is called ‘Endpoint’. One LayerZero Endpoint exists on each (supported) chain, and any chain with a LayerZero Endpoint can conduct cross-chain transactions involving any other chain with a LayerZero Endpoint. In essence, this creates a fully connected network where every node has a direct connection to every other node.
Process:
- Request a transaction to the Endpoint on chain A.
- Endpoint sends the message to Relayer and Oracle. (Relayer and Oracle are off-chain and must be independent).
- Oracle confirms the protocol header. (Oracle is Chainlink in the testnet).
- Relayer confirms the transaction request on Chain A.
- Endpoint on chain B is informed.
- Chain B processes the remaining transaction steps just as the operations in the individual blockchain.
Features:
-
High Reusability & Low Complexity
Light weight with low complexity and low deployment cost.
-
Middle Security
Even no intermediate consensus layers is alleged, it still relies on the oracle and relayer not to collude, and oracle like ChainLink itself is not completely secure.
-
Low Latency
Direct point to point communication between chains.
-
High Overhead
Chainlink calling is expensive. 1 chainlink call to ETH is 1-3 LINK.
Ecosystem:
LayerZero is live on Ethereum, Avalanche, Polygon, BNB Chain, Fantom, Arbitrum and Optimism. In 6–8 weeks consider Solana, Terra, Cosmos Hub and Osmosis slam dunks and LayerZero to be everywhere in 4–6 months, according to the CTO at LayerZero Labs.
Stargate, which is the bridge project, and the first project that launches on top of LayerZero protocol (and all 7 chains), claims to be the first bridge to solve the Bridging Trilemma.
Stargate is based on layerzero, but the current function only includes the cross-chain of stablecoins and $STG. In the future, it is likely to continue to build DEX, lending and other dapps based on LayerZero.
According to the data on the official website, since the launch of Stargate finance for 7 days, the TVL has exceeded $2 billion, and the current amount is $220 million, ranking 16th among all Defi protocols.
Axelar
Axelar stack provides a decentralized network, protocols, tools, and APIs that allow simple cross-chain communication.
There are two ways to interact with the network:
- Come in and build on top of the actual network directly in order to help you to facilitate the cross-chain flows.
- Build the application in your favorite environment in your favorite ecosystem. Interact through API directly with the gateways on the source chain where the information originated from. (just like a router which composes you to other ecosystems)
Principles:
Cross-Chain Gateway Protocol （CGP）
Connect blockchain ecosystems and routing across them. By deploying gateways in different chain ecosystems, routing between blockchains with different technical architectures is realized without changes to the chain itself.
Cross-Chain Transfer Protocol（CTP）
Application developers can deploy their dapps on any chain to perform cross-chain requests, and interact with dapps on any chain using simple API calls analogous to HTTP requests.
Process:
Pre-built Layer:
- To bridge different chains, threshold accounts(like routers) are created on each chain.
- Axelar validators create keys per user request to control threshold accounts.
During Transaction:
- The user on chain S post a request to a bridge account (or directly to the Axelar chain)
- Axelar validators run node software to pick up the request, and report to the Axelar chain.
- Once certain weighted validators report the same answer, Axelar asks validators to sign.
- The signature is included in a new block, anyone can take the signed request value from this block and post it to chain D.
- The request has been serviced. Any application on D may now take the signed value from bridge accounts, and verify the signature corresponds to the keys.
The white paper abstract it into a more understandable sentence:
An Dapps on chain A, can update its state if some other application on chain B satisfies some criteria, and perform general cross-chain requests between apps across chains.
In other words, Axelar builds 'router' on each chain and a network to achieve state transfer between chains. We could see that Axelar‘s ridging solution is also based on the most rudimentary elements of blockchain, the event and the state.
Features:
-
High Reusability
Only a gateway is needed to deploy on each chain. the chain itself does not need to make changes.
-
Middle Complexity & Middle Security
Mechanisms complexity is further increased in order to maximum decentralization. BFT Byzantine consensus is used to build a third-party chain, which is still based on the ‘validator set’ mechanism just like other ‘Trusted Bridge’. Even though mechanisms complexity is further increased in order to maximum decentralization, the security concern could not be ignored.
-
Low Latency
Semi-direct point to point communication between chains.
-
Low Overhead
Do not depend on outside oracle compared with LayerZero. Low cost for dapps.
Vision - From Official Propaganda:
For blockchain platform builders: Ability to easily plug-in their blockchains to all other blockchain ecosystems. Only a threshold account needs to be set up on the chain to plug into the network.
For dapps builders: Application builders can host their dapps anywhere, lock, unlock, transfer assets, and communicate with applications on any other chain via CTP API.
For users: Users can interact with all applications across the ecosystem directly from their wallets.
Ecosystem:
Axelar is live on Cosmos, Ethereum, Osmosis, Avalanche, Moonbeam, Fantom, e-Money, Terra, and Polygon. Use cases including NFT+Defi, universal AMM, cross-chain synthetics, and open gaming.
Axelar is also in partnership with Polkadot, Keplr, Injective, Metrika, Pangolin, Solarbeam, SpookySwap, StellaSwap, and Trader joe.
Axelar has not yet born a native app, which means opportunities. According to the latest news on the official website, Axelar announces $2.3M in initiating a multichain grant program, which has a significant title ‘Build the Web3 Super App’. The claim is as follows:
‘Build a cross-chain native application, using Axelar network to give it the distribution of the entire ecosystem. Use your imagination: we want to see Dapps that Web3 has never seen before. Integrate assets, functionality and the ability to interact with users of another dApp on another chain’.
We could see the potential and opportunities behind Axelar.
Comparison
Axelar and LayerZero represent two types of cross-chain bridge.
Axelar represents a cross-chain protocol that requires a third-party chain and an intermediate consensus layer. However, the reliance on intermediate consensus layer and token brings additional operations and delays. Compared to other chains of the same type, Axelar minimizes the loss by designing the mechanism at a more fundamental level, which allows more direct communication between chains.
While LayerZero narratively ‘removes’ the intermediate consensus layer to allow the direct communication between chains. However, LayerZero needs to rely on oracle and relayer. Technically, LayerZero still needs the intermediate consensus layer but in the form of oracle rather than an individual chain. For a long time to come, the oracle will exclusively be Chainlink. The calling cost for developers would be a big concern.
Both LayerZero and Axelar require some sort of ‘consensus mechanism’. The difference is that LayerZero adopts outside oracle to be the part of validators, while Axelar acts directly as an individual blockchain as a consensus layer.
Both LayerZero and Axelar provide plug and play development package for direct API interaction with the ‘Endpoint/gateway’ on the chain. However, compared with LayerZero which is more like just a ‘tool’, Axelar provides not only a ‘tool’, but a ‘network’ with a much bigger vision. Axelar narratively has the potential to become Trans-ecological ecology.
Besides ‘bridging ability’, ‘narrative ability’ is also an important philosophy for each cross-chain solution teams to master. Axelar theoretically has a potential to be an ecosystem of ecosystems. And we have seen that the Axelar team has finally begun to take actions, such as the official grant program that we are following: ‘The Axelerator Multichain Grant Challenge: Build the Web3 Super App’. After all, the real test for the cross-chain solutions is the ability to let the chains and apps adopt it. We believe that both LayerZero and Axelar are worth looking forward to.
Finally, add LayerZero and Axelar to the existing trade-off comparisons:
*Welcome to communicate, criticize and correct. Write to Mike Ma at mike@capitalant.com.
References:
-
’What Does the Ideal Cross-Chain Bridge Look Like?‘
https://medium.com/a-t-capital/what-does-the-ideal-cross-chain-bridge-look-like-a7c1113f9a18
-
‘James Prestwich : Building Bridges, Not Walled Gardens’
-
‘Why are the Bridges Burning? James Prestwich’
-
‘Axelar Network: Connecting Applications with Blockchain Ecosystems’
https://axelar.network/wp-content/uploads/2021/07/axelar_whitepaper.pdf
-
‘LayerZero: Trustless Omnichain Interoperability Protocol’
https://layerzero.network/pdf/LayerZero_Whitepaper_Release.pdf
-
‘BlockChain Principle, Type & Application & Why You Should Care About It?’
-
‘Cross-chain final battle’