After the Ethereum merger failed to improve scalability, EIP-4844 is the killer weapon for Ethereum expansion.

All eyes may be on Ethereum’s upcoming transition to Proof-of-Stake (PoS) consensus, but there’s a lesser-known upgrade that will reduce Ethereum’s transaction costs and improve scalability, separate from what’s being called The Merge.

The EIP-4844 upgrade will introduce “proto-danksharding” in the next Ethereum fork upgrade after The Merge, which is expected to increase the scalability of Layer 2 Rollups by up to 100 times while paving the way for full sharding.

This is important because Ethereum’s current roadmap is providing banking services for the network’s thriving layer 2 ecosystem, which will handle scalability and fee mitigation until sharding is further implemented. Sharding will spread Ethereum’s computational load across an ecosystem of small chains working in parallel, and it is expected that layer 2 chains will become shards of the network in the future.

On Ethereum’s leading rollup network, a simple token transfer costs about $0.1, while a swap is estimated at about $0.15, according to L2fees. But that’s not good enough, with Ethereum co-founder Vitalik Buterin tweeting in May that L2 transaction fees must be below $0.05 “to be truly acceptable.”

The Merge, which will be launched in mid-September, will reduce the energy consumption of the Ethereum network by more than 99% and the issuance of new ETH by about 90%. However, many in the community mistakenly expect that this upgrade will significantly improve Ethereum's scalability and reduce network fees.

Tim Beiko, protocol development coordinator at the Ethereum Foundation, told The Defiant that proto-danksharding is expected to go live six to nine months after The Merge. However, he noted that “upgrade timelines are notoriously difficult to predict” and that EIP-4844 will need to pass governance before it can be implemented.

Sharding and Proto-Danksharding

Proto-danksharding is named after Dankrad Feist, a member of the Ethereum Foundation. Feist designed danksharding, the current version of sharding that is planned to be introduced as part of the Ethereum scaling roadmap.

Proto-danksharding is the first of a two-part process to introduce sharding to Ethereum. It will implement much of the infrastructure for sharding before the network actually shards. This will include the transaction format, validation rules, consensus and execution logic, and gas price adjustments contained in the Danksharding specification.

Once implemented in two to three years, full danksharding is also expected to further increase Ethereum’s scalability by an order of magnitude.

With danksharding, each validator will only have to download a small fraction of the blocks, increasing block size and throughput in rollups by approximately 100x compared to today.

Rollups have emerged as Ethereum’s leading layer 2 scaling solution and are implemented by companies like Arbitrum, Optimism, and Polygon. They work by bundling together transactions that are executed on a low-cost layer 2 network, which are then submitted in batches for verification on Ethereum’s base layer to significantly reduce transaction fees.

Mofi Taiwo, an engineer at Optimism, the team behind Optimism, told The Defiant that “from the perspective of L2 and other applications, the user experience of proto-danksharding is the same as full sharding.”

“This allows Ethereum to provide a complete data availability solution for rollups while developing full sharding, significantly reducing costs,” he added.

Replace Calldata with Blob

Buterin co-authored the EIP-4844 proposal with other ethereum researchers in late February, about a week after the first proto-danksharding prototype was developed at the ETHDenver hackathon.

Proto-danksharding will allow transactions to carry "blobs" instead of calldata. Blobs are smaller and cheaper for the network than the calldata they will replace.

Ethereum Foundation and Optimism researcher Diederik Loerakker tweeted that “data blobs are the first milestone towards full Ethereum sharding, enabling rollup… capacity to grow 100x.”

Buterin wrote that the upgrade would bring “huge scalability gains, as this data would not compete with the gas usage of existing Ethereum transactions.”

Delete historical data

Once EIP-4844 goes into effect, historical data will also be deleted from the network after 30 days. “The purpose of the Ethereum consensus protocol is not to guarantee the storage of all historical data forever,” Buterin wrote. “Rather, the purpose is to provide a highly secure real-time bulletin board and leave room for other decentralized protocols to perform longer-term storage.”

The elimination of the need to store historical data is one of the key features of EIP-4844, which will drive the scalability of rollups.

Taiwo explained,

“Rollups…generate a lot of data, and they need to send that data back to L1, and proto-danksharding gives them a cheaper way to do that. This design is particularly well suited for rollups because they don’t require permanent storage, but rather a strong guarantee that the data has been available on the Ethereum network for some period of time.”

Buterin said erasing historical data would only pose a risk to individual applications, not the Ethereum protocol itself. “It makes sense for applications to bear the burden of storing data relevant to themselves. Block explorers, API providers, and other data services may store the full history.”

In a recent appearance on The Defiant podcast, Ethereum Foundation researcher Justin Drake described the need for every validator to download the entire Ethereum blockchain as “redundant work.”

Looking ahead

Taiwo pointed out that the number of Rollups using EIP-4844 will affect the scalability gains achieved when proto-danksharding goes online.

He said,

“More Rollups leveraging EIP-4844 will increase competition for blob space on the network, but fees will still be significantly lower than a network without the scaling advantages of proto-danksharding.”