In the world of cryptocurrencies and blockchain technology, Ethereum is one of the most important platforms. Recently, there has been a debate about increasing the gas limits on Ethereum's Layer 1 (L1), even when most transactions are handled on Layer 2 (L2) solutions. This article explores why it might be useful to increase these gas limits, even in a world where most applications and transactions are managed on L2.
Censorship Resistance
One of the main advantages of blockchains is censorship resistance. This means that if a transaction is valid and you have enough funds to pay the fee, you should be able to include that transaction in the blockchain quickly. In some cases, such as with decentralized finance (DeFi) protocols, even a five-minute delay could be critical.
Ethereum's Layer 1 is highly decentralized, making it very difficult to censor a transaction for more than a few blocks. However, Layer 2 solutions depend on a more concentrated set of block producers or a centralized sequencer, which could choose to censor users. Some L2s have a forced-inclusion mechanism that allows users to send transactions directly through Layer 1, but this depends on Layer 1 fees being low and having enough space for users to send transactions even if an L2 censors many users.
Moving Assets Between L2s
Often, users need to move assets from one L2 to another. For high-volume assets, the most practical method is through exchange protocols like ERC-7683. However, for low-volume assets or NFTs, this is not possible, and users must send transactions through Layer 1.
Today, a withdrawal costs about 250,000 Layer 1 gas and a deposit another 120,000 Layer 1 gas. With an ideal design, this flow could be optimized significantly. For example, moving an NFT from one L2 to another could cost only 7,500 Layer 1 gas. This implies a need to scale Layer 1 by about 5.5 times to reach the goal of a $0.05 transaction cost.
L2 Mass Exits
Another important property of L2s is the ability to exit to Layer 1 if the L2 fails. In Plasma systems, it may be necessary to exit within a week if data becomes unavailable. Even in optimistic rollups, a hostile governance upgrade could give users a 30-day timeline to withdraw their assets.
If a single Plasma chain fails and an exit costs 120,000 gas, how many users could exit within a week? With Ethereum's current capacity, only 7.56 million users could exit. If more efficient mass-exit protocols were implemented, this number could increase significantly. However, to support a mass exit of users, Layer 1 would need to scale by about 9 times.
Issuing ERC20s on L1
Many tokens are being launched on L2s today. This has an underrated security concern: if an L2 undergoes a hostile governance upgrade, an ERC20 launched on that L2 could start issuing an unlimited number of new tokens, and there would be no way to stop those tokens from leaking into the rest of the ecosystem. If a token is issued on Layer 1, the consequences of one L2 going astray are mostly limited to that L2.
To make issuing ERC20s on Layer 1 a popular option, it needs to be cheap. Currently, the cost of launching an ERC20 token on Layer 1 is about $61.76. With an ideal design, this cost could be reduced to $4.5, but to reach the goal of $0.25, Layer 1 would need to scale by about 18 times.
Keystore Wallet Operations
Keystore wallets are a type of wallet that has modifiable verification logic (for changing keys, signature algorithms, etc.) that automatically propagates across all L2s. The verification logic sits on Layer 1, and L2s use synchronous reads to access the logic. If each user needs to change a key or upgrade an account once a year, and there are 3.1 billion users, this would require a gas capacity of about 3.3 times the current target.
With very optimized designs, key change operations could be reduced to 7,500 gas, which would allow keystore updates to be made with about half of Ethereum's current gas capacity.
L2 Proof Submission
For cross-L2 interoperability to be fast, general-purpose, and trustless, L2s need to post to Layer 1 frequently, so they can be directly aware of each other's state. With current technology (ZK-SNARKs), this has a cost of about 500,000 gas per L2, limiting Ethereum to supporting only 36 L2s. Moreover, this is economically unviable: the annual cost of submitting would be about $49 million per L2.
With an ideal design, this cost could be reduced to 10,000 gas per submission, making the cost of submitting to Layer 1 every slot about $1 million per year per L2. To make submitting to Layer 1 every slot a no-brainer, Layer 1 would need to scale significantly.
Conclusion
This analysis shows that there is significant value in scaling Layer 1 gas limits by about 10 times, even in a world dominated by Layer 2 solutions. This implies that short-term Layer 1 scaling that can be done in the next 1-2 years is valuable, regardless of what the long-term picture ends up looking like.
https://vitalik.eth.limo/general/2025/02/14/l1scaling.html
