Discuss Practical Solutions for Improving Ethereum Transaction Confirmation Times

An important aspect of blockchain user experience is transaction confirmation time. In recent years, Ethereum has made significant progress in this regard. Currently, transactions sent by users on L1 can typically be confirmed within 5-20 seconds, comparable to the experience of paying with a credit card. However, it is still valuable to further shorten confirmation times, as certain applications may even require millisecond-level latency. This article will explore some feasible options for Ethereum in this area.

Overview of Existing Solutions

single-slot finality

Currently, Ethereum's Gasper consensus uses a slot and epoch architecture. A slot occurs every 12 seconds, during which validators vote on the chain head, and all validators have the opportunity to vote once within 32 slots (6.4 minutes). These votes are interpreted as messages similar to the PBFT consensus algorithm, providing strong economic guarantees of finality after two epochs (12.8 minutes).

However, this method has issues with complexity and excessive time. Single Slot Finality (SSF) replaces this architecture with a mechanism similar to Tendermint, allowing block N to be finalized before block N+1 is generated. SSF retains the "non-active leakage" mechanism, allowing the chain to continue operating even when more than 1/3 of the validators are offline.

The main challenge of SSF is that each staker needs to publish two messages every 12 seconds, which places a significant burden on the chain. Although there are some mitigation solutions, such as the Orbit SSF proposal, users still need to wait 5-20 seconds.

Rollup Pre-Confirmation

Ethereum adopts a rollup-centric roadmap, where L1 provides data availability and other functionalities, while L2 protocols build upon this to offer users larger-scale services. This leads to a separation of concerns: L1 focuses on censorship resistance, reliability, and core functionalities, while L2 directly addresses user needs.

In theory, L2 can create its own "decentralized sequencer" network that signs blocks every few hundred milliseconds. However, this requires L2 to do nearly the same amount of work as creating a new L1, resulting in slow progress. Therefore, some have proposed allowing all L2s to share a pre-confirmation mechanism within the Ethereum scope: the base pre-confirmation.

Basic Pre-confirmation

The basic pre-confirmation leverages the complexity of Ethereum proposers to incentivize them to provide pre-confirmation services. Users can pay an additional fee to receive an instant guarantee that their transaction will be included in the next block. If the proposer violates the commitment, they will face penalties. This mechanism is applicable not only to L1 transactions but also to L2 blocks based on "rollups."

Possible Directions for the Future

Assuming single-slot finality is achieved, with technology similar to Orbit to reduce the number of validators signing each slot while lowering the staking threshold. The slot duration may increase to 16 seconds, combined with rollup pre-confirmation or base pre-confirmation to provide users with faster confirmation. This essentially forms an epoch-slot architecture.

The reason this architecture is hard to avoid is that achieving a rough consensus takes less time than achieving the maximum level of "economic finality." Factors include the number of nodes and the "quality" of the nodes. If we can rely on a specialized subset of nodes to reach an approximate agreement, while using the full validator set to determine finality, it may reduce the confirmation time to about 2 seconds.

Therefore, exploring the design space of epoch-slot architecture with stronger attention separation is valuable.

Strategy Selection for L2

Currently, there are three reasonable strategies for L2:

1. Technically and spiritually "based on" Ethereum, optimizing its technical attributes and values.
2. Become a "server with blockchain scaffolding," combining server efficiency and blockchain security.
3. Compromise method: Fast chains combined with Ethereum provide additional interoperability and security.

For certain applications, a 12-second block time is sufficient. For others, the only solution is the epoch-slot architecture. The key question is how well Ethereum's native epoch-slot architecture can perform, as this will affect the significance of other solutions.

Currently, we are still far from the final answers to these questions. The complexity of block proposers remains uncertain. New designs like Orbit SSF provide space to explore more possibilities. The more options we have, the better we can serve L1 and L2 users and simplify the work of L2 developers.