Ethereum L1 zkEVM Exploration

There is a common misconception in the industry that Ethereum can only be scaled through Layer 2 (L2) networks. This view may partly stem from the past overemphasis on L2 Rollups, neglecting the L1 scaling roadmap. With the maturity and widespread adoption of L2 technology, it is now time to reevaluate the possibilities of L1 scaling.

First of all, it needs to be clarified that ETH, as a cryptocurrency asset, has its expansion paths far beyond L2. In fact, ETH can be expanded in multiple ways, including L1, sidechains, other public chains, L2, quasi-L2 structures (such as validium and optimistic chains), and even centralized exchanges and service providers. Currently, millions of ETH have been bridged to various chains and non-blockchain environments. While L1 and mature L2 provide native security guarantees for ETH, other solutions, despite possibly having different security assumptions, can still expand ETH or Ethereum assets.

It is worth noting that the value of an asset does not solely depend on its method of expansion. For example, Bitcoin, despite being primarily expanded through centralized services, remains the dominant asset in the industry. This indicates that for an asset to gain value, the key lies in obtaining recognition from the wealthiest groups, families, and institutions.

Returning to the topic of L1 zkEVM upgrades. The use of Zero-Knowledge Proofs (ZKP) to scale blockchains is a long-standing concept, predating the birth of Ethereum. With the feasibility proven by ZK Rollups through projects like Loopring, StarkEx, and zkSync in early 2020, research on ZK-SNARKing Ethereum entered the fast lane. In 2021, the term "zkEVM" began to circulate widely.

The implementation path for the L1 zkEVM upgrade may include the following steps:

1. Test Type-2/2.5 and Type-1 zkEVM Rollup in production environments through upcoming projects like Scroll, Linea, and Taiko.

2. Implement prerequisites such as EIP-4844, statelessness, and PBS (Proposer-Builder Separation).

3. Deploy an embedded zkEVM bridge that allows Type-1 zkEVM to be deployed on top of L1 for real-world testing of the code and zk circuits intended for L1 zkEVM.

4. The L1 execution layer is ready for the zkEVM upgrade. Builders sort transactions, generate proofs, and submit them to the consensus layer.

5. In the consensus layer verification proof, non-builder nodes only need to verify these proofs, thereby efficiently validating a large number of transactions, including L2, L3, etc.

6. Keep the embedded zkEVM bridge on top of the L1 execution layer, or move it to the consensus layer.

It is worth noting that embedded L2 Rollups have their own trade-offs. The initial zkEVM may be slower, with limited throughput and functionality, and the upgrade cycle may be longer. They will be fully operated by Ethereum node operators, with no independent governance. Therefore, innovation will still mainly occur on traditional L2s, and it is expected that most users will continue to use traditional L2.

In the future, traditional L2, hybrid versions of L2 (such as validium or optimistic chains), embedded L2, and embedded L1 Rollups will provide users with different trade-offs and functional choices to collectively meet the diverse needs of the blockchain ecosystem.

Of course, all of this may be unnecessary. We may not actually need such a high throughput, and a more prudent approach might be to maintain the status quo of L1. In any case, this decade is coming to an end, and the future development remains full of uncertainties. However, the vision of this L1 zkEVM is undoubtedly exciting and worth our continued exploration and anticipation.