#热门话题 #Ethereum(ETH) #EIP4844(Proto-Danksharding): What impact will it have on L2s transaction costs?

Which L2 will win the fee war?

⚔️💰

🧵⬇️

Let’s dive deeper into analysis and predictions + TLDR add-on

A major Ethereum upgrade since the merger is underway: the Dencun hard fork bringing EIP-4844 (Proto-Danksharding) will be released on March 13 and aims to make L2 transactions cheaper.

What will be the impact? Which L2 will win the fee war?

Methodology: 1) Involves evaluating transaction costs for various L2s (Starknet, ZkSync, Optimism, Scroll and Arbitrum), comparing average DA to execution costs (%) 2) Then projecting these costs onto new DA costs after 4844 3 scenarios (pessimistic, pessimistic, medium and DA=0) 3/16

This analysis includes all types of transactions (transfers, exchanges, multiple calls...)

(Collecting the exact same and consistent data source in each L2 is very challenging. Therefore, we will consider the data that each L2 has access to.)

Now, let's get started

🏁

- @Starknet

The accompanying chart shows that DA costs currently account for approximately 90% of total costs, with execution costs accounting for the remaining 10%.

The green and red sections show the average DA cost on L1 (pre-EIP-4844), and the purple bar represents the execution cost (expressed as a percentage).

- @ZkSync

The attached graph shows that the average DA cost is about 65% and the execution cost is 35%. -Although there are some meaningful peaks-.

- @Scroll_ZKP

Scroll exp ( http://scrollscan.com ) shows an average DA cost of 62-63% for swaps/multiple calls and an average cost of 90% for transfers. We estimate the average DA cost to be 70%.

Without weighting each type of transaction, we would consider the average DA cost to be 70% (compared to 30% for exec)

- @Optimism (OP stack)

According to their March 3rd article, the OP claimed that transaction fees after 4844 were reduced by 70x according to their "Super Chain Savings Estimator" (SSE).

However, the exact method used to justify this is not specified in detail.

We must now rely on this data and wait 4844

8/16

- @arbitrum

Unfortunately, data were not available to compare DA weights with execution costs, so we were unable to include them in this analysis.

Offchain Labs published a great article about it https://medium.com/offchainlabs/eip-4844-what-does-it-mean-for-l2-users-5e86ebc4c028…

Now that we have the average cost of DA versus the execution cost of L2, let’s consider three post-EIP-4844 scenarios to measure the impact on each L2 price

1) Pessimistic scenario

2) Medium scene

3) Extreme scenario: DA = 0

1) If DA costs drop to 50%, overall transaction fees will decrease:

- STARKNET: x1. 8 (overall costs 55% of current levels)

- SCROLL: x1.5 (overall cost is 65% of current level)

- Zksync: x1.5 (overall cost is 67.5% of current levels)

- ARB: Not applicable

2) If DA cost drops to 90%, overall transaction fees will decrease:

- Starknet: x5 (overall cost is 19% of current level)

- SRCOLL: x2.7 (overall cost is 37% of current level)

- Zksync: x2.4 (overall cost 41.5% of current levels)

- ARB: Not applicable

in conclusion

🏆

If Optimism's predictions hold true, they'll win the 4844 fee war

But the details are few and we have to rely on them

They claim 70x reduction in Base, Zora, Mode fees

This means that in the event DA drops to 0, the DA cost of the OP stack will be 98.5% of the transaction fee

let's see!

Conclusion (2):

The remaining rankings are as follows:

- STARKNET

🥇

- SCROLL

🥈

- Zksync

🥉

- ARB: Not applicable

Now let’s wait until March 13th when EIP-4844 is released

✨🫡