Next, EIP-4844

tags: thread

The next big goal of Ethereum after the merge is sharding (Danksharding)

And because merge and sharding (including other upgrades of Ethereum) work in parallel, sharding has made a lot of progress

Danksharding

Danksharding is a new sharding design proposed by @dankrad that focuses on data availability rather than code execution

This is a long-term upgrade goal of Ethereum, and its purpose is to further reduce the cost of Ethereum as a data settlement layer

Rollups as a mid/long-term scaling solution for Ethereum, the cost will be further reduced after Danksharding

EIP-4488 and EIP-4844

For the short term, there are mainly two proposals, namely EIP-4488 and EIP-4844👇

• EIP-4488 reduces the cost of calldata from 16 gas per byte to 3 gas per byte and limits the maximum calldata per block
• EIP-4844 (Proto-Danksharding) proposes a new transaction type (blob-carrying transaction)

Its role is similar to calldata, returning rollup batch data to Layer 1 for data availability, but because it is specially designed for rollups, it's larger and cheaper

EIP-4844: Proto-Danksharding

Anyways, Danksharding and EIP-4844 are very interesting, and it mainly works with the consensus layer (i.e. Beacon Node), not the execution layer

Blob is stored in Beacon Node, and future sharding is coordinated through Beacon Node too

So it is very suitable to go deeper after understanding "the merge" (PoS and Beacon Chain)

Because you need to be familiar with the post-merge architecture of Ethereum, Beacon Node's workflow, the validator mechanism, etc.

EIP-4844 has made a lot of progress so far👇

• For example, the ongoing KZG ceremony requires only one honest participant to complete the trusted setup. Public contributions are expected to open in November 2022 (after @EFDevcon), when everyone can join the ceremony!
• As well as the devnet that has been launched in recent days, users can build geth+prysm nodes (request ETH for devnet), build blobs, and upload them

Conclusions

Eventually Danksharding will introduce data sharding based on EIP-4844, which is a more distant future

In the end, the gas cost of rollups will be reduced by more than 100x, not to mention that we also have solutions such as Validium where DA is not on Ethereum (very suitable for applications that do not require all data to be on the chain)

Thread