Introduction
The Merge, a pivotal milestone for Ethereum, was completed on September 15. According to Vitalik Buterin's Ethereum protocol development roadmap released in late 2021, the next critical phase is The Surge—addressing Ethereum’s scalability issues to reduce transaction fees and increase throughput. The Surge revolves around a rollup-centric roadmap, enhancing the scalability of L2 rollups while inheriting Ethereum’s network security.
This article explores a key component of this technical roadmap: EIP-4844 Proto-danksharding, which significantly reduces the cost of data storage for rollups and increases data capacity. EIP-4844 upgrades the Ethereum network by introducing a new transaction type that carries short-lived blob data—a cheaper alternative to the current calldata method. Moreover, EIP-4844 lays the groundwork for the full implementation of Danksharding, which promises an additional 10–100x scalability boost.
Ethereum Sharding Tech Roadmap
Protolambda, a former Ethereum Foundation researcher, summarizes the current state of Ethereum’s sharding design:
Executable "shard chains" with crosslinks have been phased out in favor of: implementing EVM within the beacon chain; adopting a rollup-centric roadmap using Data Availability Sampling (DAS) to scale Ethereum’s base layer without increasing complexity.
This simplification stems from two key reasons:
- Avoiding additional L1 complexity (sharding specs were rewritten multiple times, delaying implementation).
- Leveraging modular blockchain architecture, where L1 serves as a data availability layer for rollups, offloading computation to L2 execution layers.
Modular Architecture in Ethereum
Layered Approach
- Pre-Merge: Monolithic PoW chain.
- Post-Merge: L1 consensus layer (PoS) + L1 execution layer (EVM), connected via the Engine API (Ethereum’s first modular design).
- L1-L2 Connection: An API bridges L1 (data/security layer) and L2 (execution layer), with fraud/validity proofs ensuring trustlessness.
Data Availability: The Scalability Bottleneck
Rollup costs are dominated by data availability fees (publishing data to L1 via calldata). Current solutions like calldata compression and EIP-4488 (reducing calldata cost from 16 to 3 gas per byte) are stopgaps. EIP-4844 offers a sustainable alternative.
EIP-4844 Proto-danksharding
Key Features
Blob-carrying Transactions:
- New transaction type carrying blob data (for L2).
- Blobs are cheaper than calldata (not read by contracts) and pruned after ~1 month, mitigating state bloat.
Sidecar Model:
- Blob data is separated from transactions (stored in consensus-layer nodes) and deleted post-use.
- Ensures data is available long enough for state reconstruction/challenges.
Workflow:
- Sequencer submits blob data → L1 finalizes data.
- Blob sidecar splits from transaction → Execution occurs in Execution Payload.
- L2 validators download sidecars to sync state.
Conclusion
EIP-4844 addresses the data availability bottleneck for rollups, enabling exponential scalability. By introducing blob transactions, Ethereum’s base layer becomes a cost-efficient data repository for L2s without compromising security. As the precursor to Danksharding, Proto-danksharding paves the way for Ethereum’s next 10–100x leap in throughput.
FAQs
Q1: How does EIP-4844 reduce rollup costs?
A1: By replacing expensive calldata with blob transactions—unread by contracts and pruned post-use, cutting fees 10–100x.
Q2: What’s the difference between blobs and calldata?
A2: Calldata lives in execution payloads (permanent), while blobs are stored in consensus-layer nodes (temporary, deleted after ~1 month).
Q3: Is EIP-4844 compatible with future Danksharding?
A3: Yes. Proto-danksharding is a stepping stone, introducing the infrastructure needed for full Danksharding’s massive scalability boost.
👉 Learn more about Ethereum’s modular future
👉 Explore rollup-centric roadmaps