What does EIP 4844 mean for Ethereum Rollup?

Author: Bessie Liu, Blockworks; Compiler: Deng Tong, Golden Finance

The Dencun hard fork is expected to be launched in March 2024, and multiple Ethereum Improvement Proposals (EIPs) will be launched on the Ethereum main Online.

Among them, EIP-4844 (also known as proto-danksharding) has been receiving special attention.

In today’s Ethereum ecosystem, Layer 2 scaling solutions such as Rollups play a vital role in increasing transaction throughput and reducing costs. These rollups "inherit" Ethereum's security by having execution layer nodes store temporary copies of the network's ledger and history. This process verifies the accuracy of transactions.

However, the current method of achieving this (by writing the data to "calldata") is expensive, costing around $1,000 per megabyte. Due to the significant costs involved, this pricing model poses significant challenges for Layer 2 solutions, especially during periods of high demand.

EIP-4844 will introduce blob transactions as an alternative to calldata. These blob transactions are a temporary data storage mechanism designed to meet the data availability needs of Ethereum Rollup solutions. Blobs are essentially large packets of data that can be processed and stored more efficiently than current methods.

Offchain Labs co-founder Edward Felten explained in an interview that 4844 will provide a lower level of data availability services but still meet the needs of Rollups.

"It is not available for Ethereum's execution layer and is retained by Ethereum nodes for only 18 days," Felten said. "This is sufficient for Rollup."

Felten noted that data is currently used in two ways. One of them is to prove the result that happens when Rollup.

In the meantime, it is important to ensure that the data itself is available to everyone so that different actors can participate in ensuring that transactions are accurately recorded.

"The first 18 days are long enough for any realistic proof-of-concept mechanism to do its job," he said.

Felten said that after the 18-day attestation period, users on that particular Rollup may still be concerned about their Rollup's data, but that becomes the Rollup's responsibility to provide and track the information.

Do Blobs make transactions cheaper?

Blob pricing is designed so that when usage is high, the price will increase, and when usage is low, the price will decrease.

Felten pointed out that it is difficult to accurately estimate how much 4844 will reduce the cost of publishing data for Rollup solutions due to two unknown factors.

“One thing that might happen if transactions become cheaper is that users will make more transactions ,"Felten said. "We really don't know how much Layer 2 traffic will increase and how much more transactions will occur as transactions become cheaper."

This means, If transaction prices become cheaper, it may create more demand for data publishing and blob space, which in turn will push the price of blobs higher.

Another unknown factor is how Rollup itself will use these blobs.

Felten said: “Rollup tries to optimize cost and ultimate latency for users, and we believe that if Blobs are more expensive low, we will release smaller blobs more frequently." "If the price of blobs drops significantly, this will cause Rollup to use more blobs, which will tend to put upward pressure on the price."

Felten noted that it is important for Rollup to consider exactly how long it should wait before batch-posting data to Ethereum.

Currently, on Arbitrum with calldata, the network accumulates approximately 100-120 KB of data in batches before publishing it to Ethereum. But Felten explained that unlike calldata, where Layer 2 must pay per byte, with blobs, rollup must purchase the entire blob, whether it is used or not.

"That means the tradeoff is a little different, because if you only have half a blob of data and you publish it, you're buying the whole blob but only using half of it," Felten said.

"If you wait longer, you will publish more data in the blob, thus saving costs, but on the other hand, your user transactions will take longer to reach Ethereum, so they The final result will be reached later,” Felten added.

Data publishing strategy

One potential way to reduce the cost of publishing data is through shared publishing. This is where two different Rollups might choose to merge their data and publish it to the Ethereum mainnet.

Offchain Labs researcher Akaki Mamageishvili pointed out that If Rollups wants to share data releases to save costs, then they must solve One issue is determining cost allocation.

"A faster rollup or a busier rollup will save less latency costs, while a less busy rollup or a slower rollup will save more latency costs, so cost sharing should not be proportional,” Mamageishvili said.

He added that rollups need to determine which other Layer 2s they want to "partner" with, noting that larger rollups may not need to work with others, while smaller rollups may need to find others. Smaller rollups to spread costs and improve latency.

Data availability options

Ethereum itself is not the only blockchain that can publish data. Data availability solutions such as Celestia and Eigenlayer also hope to solve the data publishing problem through data availability sampling.

Each Layer 2 solution has the ability to choose exactly where they want to publish their data. In the case of Arbitrum One and Arbitrum Nova, a DAO vote is currently underway to ensure that the Arbitrum chain will have the ability to leverage EIP-4844 immediately following the Dencun upgrade.

"In my opinion, it's a decision that each chain makes between cost and trust, because you trust the data availability system you're using. Each chain can make its own decision , I don’t think it’s one size fits all.”