Stargate Integration with Espresso Network

General
1

Summary

Request for Stargate to integrate with the Espresso Network to enable bridging from supported chains within seconds. Espresso provides reorg protection to rollups so that applications like Stargate can know within seconds that a bridging transaction is guaranteed to eventually be finalized on the L1, without having to trust the centralized sequencer for soft-confirmations. Currently Rari Chain is the first chain that could benefit from faster confirmations bringing routes from 1 minute down to a couple of seconds.

Background

The Espresso Network consists of a decentralized validator set running the HotShot consensus protocol. HotShot is byzantine-fault tolerant, meaning that as long as two thirds of the validator set is honest, finalized blocks can never be reverted. HotShot achieves this finality within just a few seconds.

Rollups that integrate with the Espresso Network require HotShot to finalize all of their transaction batches. Additionally, the rollup settlement contract is modified such that it will only settle batches that were finalized by HotShot.

An example of a team that is using the Espresso Network would be Rari Chain. They are live on mainnet using Espresso Network’s Hotshot for secure and fast confirmations. With the Rari Chain integration they still sequence their own transactions but publishes blocks to Espresso allowing for Espresso to confirm them.

This ensures that Stargate can immediately fulfill bridging transactions upon confirming they are finalized by the Espresso Network, without having to worry about reorg risk or a compromised sequencer.

In practice, this should allow Stargate to safely bridge tokens within just a few seconds, simply by running a rollup node that is modified to update its state based on the finalized output of HotShot. Alternatively, Stargate can use an RPC provider who runs such a node in its place.

Espresso has built open-source implementations of these modified nodes that can be used in a plug-and-play manner, simply through running a docker container.

To learn more about Espresso Network we have documentation available on our website here

Motivation

Rollups that integrate with Espresso achieve strong reorg protection and stronger confirmations in a matter of seconds. If Stargate utilizes the Espresso Network’s finality guarantee once these chains are integrated with Espresso, these bridging times could safely be reduced from ~1 minute to below 10 seconds!

An example of routes for chains that are committed to integrate with Espresso and their speed currently on Stargate are:

  • Rari Chain: USDC, USDT - Fastest route: ~1min
  • Apechain: APE - Fastest route: ~1min
  • Superposition: USDC - Fastest route: ~1min
  • Plume: USDC - Fastest route: ~1min

These numbers above taken from the Stargate bridging UI.

Proposal

DVNs operating on Stargate will rely on the finalized output of the Espresso Network to fulfill all bridging requests originating from chains that are integrated with the Espresso Network.

Once the Espresso Network’s fast confirmations are successfully operational after 2 months supporting Rari Chain, Espresso will make a proposal to automatically opt-in/support all chains that integrate with the Espresso Network and Stargate to benefit from Espresso’s confirmations.

Current list of chains that are looking to utilize Espresso for confirmations that are also live on Stargate:

  • Superposition
  • Plume (L2)
  • Rari Chain
  • Ape Chain

Conclusion
Approval of the proposal will mean Stargate would utilize the Espresso Network for fast confirmations when bridging in and out of supported chains. This will allow Stargate to bridge out of rollups integrated with Hotshot within a few seconds whilst maintaining strong security guarantees with Stargate being protected against reorgs and a compromised sequencer.

1 Like
2

Hi Lawrence, thanks for providing us with this proposal.

I’ll be reading more on how Espresso Network works. I understand the benefits for users, but could you also cover the potential risks and how we’re mitigating them?

2 Likes
3

Hey Sendok! To quickly clarify would are you referring to:

  1. Risk for users that an espresso integration could bring about or
  2. The risks that are being mitigated when Stargate uses the Espresso network
1 Like
4

Could you help cover both the risks to users and the risks to Stargate? What could go wrong, and what are the worst-case scenarios in these situations?

2 Likes
5

A worst case scenario could arise if 34/100 nodes on the Espresso Network (also known as HotShot consensus) are malicious or become corrupt This is highly unlikely, as it would require a minimum of Espresso Systems colluding with 5 other node operators. Firstly, the node operators running the Espresso Network are some of the most well known and respected in the space (see full list here). Secondly, this risk would be mitigated once the Espresso Network completes a Proof of Stake (PoS) upgrade scheduled for Q2 2025 and will introduce slashing in a future release. This is outlined as well through our roadmap here.

If this worst case scenario occurs, the risk of loss of funds would fall on whoever is providing liquidity for Stargate. This would be due to the liquidity provider trusting the Espresso Network for confirmations prior to a chain reaching finality. For example:

  • A user makes a deposit on chain A
  • The Espresso Network confirms the deposit
  • The LP transfers to user on chain B
  • The Espresso Network reorgs, allowing chain A to finalize a state where the deposit never happened
  • The LP is unable to claim tokens/reward on chain A
2 Likes
6

Note that the core risk here is the same as it is today, namely a reorg on the source chain of a Stargate transfer causing LPs to lose funds. But as discussed, the security that prevents the source chain from reorging becomes much stronger thanks to Espresso. As there is no longer a single point of failure (the centralized sequencer) that can cause this loss of funds via a reorg to occur once Espresso confirms the transaction.

1 Like