Proto-Danksharding

Definition

Proto-Danksharding, formally known as EIP-4844, is an Ethereum network upgrade that introduces a new transaction type called “blob-carrying transactions” to provide a cheaper and more scalable way for Layer 2 rollups to post data to Ethereum Layer 1. Implemented as part of the Dencun (Deneb + Cancun) hard fork in March 2024, Proto-Danksharding serves as an intermediate step toward full Danksharding – Ethereum’s long-term data availability scaling solution. Before EIP-4844, rollups like Arbitrum, Optimism, and zkSync had to post their transaction data as calldata on Ethereum, which was expensive because calldata is permanently stored by all Ethereum nodes. Proto-Danksharding introduces “blobs” – large binary data objects (approximately 128 KB each) that are attached to blocks but are automatically pruned (deleted) after approximately 18 days. This temporary storage model dramatically reduces costs because blobs do not compete for Ethereum’s execution gas and have their own separate fee market. Each block can carry a target of 3 blobs (increased to 6 with Pectra) with a maximum of 6 (increased to 9 with Pectra). The result was immediate and dramatic: Layer 2 transaction fees dropped by 80-95% following the Dencun upgrade, making rollups significantly more affordable for end users. Proto-Danksharding also introduces the KZG commitment scheme for blob verification, laying cryptographic groundwork for full Danksharding and data availability sampling in the future.

Read Also: Render Network (RNDR)

Origin & History

Date	Event
2022	Dankrad Feist proposes Danksharding as Ethereum’s data availability scaling approach
2021	Rollup-centric roadmap established; Ethereum commits to scaling through Layer 2s
2022 Feb	Vitalik Buterin and Dankrad Feist propose EIP-4844 as a stepping stone to full Danksharding
2022 Q3	EIP-4844 development begins with new transaction type (Type 3) and blob format specification
2023 Q1	KZG ceremony completed – over 140,000 participants contribute randomness for blob commitments
2023 Q2	EIP-4844 testing begins on devnets; blob transaction format finalized
2023 Q4	Dencun upgrade (containing EIP-4844) deployed on Goerli and Sepolia testnets
2024 Mar 13	Dencun hard fork activates on Ethereum mainnet; Proto-Danksharding goes live
2024 Mar-Apr	Layer 2 fees drop 80-95%; Arbitrum, Optimism, Base see sub-cent transaction costs
2026 May	Pectra upgrade doubles blob capacity (target 6, max 9), further reducing L2 costs

“EIP-4844 is the single most impactful Ethereum upgrade for Layer 2 users – it makes rollups actually cheap, not just cheaper.”
Dankrad Feist, Ethereum Researcher

How It Works

Component	Description	Purpose
Blob	~128 KB data object attached to blocks	Stores rollup data temporarily at low cost
Blob Base Fee	EIP-1559-style dynamic fee for blob space	Creates an efficient, separate market for data availability
KZG Commitment	Cryptographic commitment to blob contents	Allows verification without downloading full blob data
Type 3 Transaction	New transaction format carrying blob references	Enables rollups to submit data via blobs instead of calldata
Blob Pruning	Automatic deletion of blobs after ~18 days	Prevents permanent state bloat while giving rollups time to process data
Versioned Hashes	References to blobs stored in the execution block	Links execution layer to blob data without embedding it in blocks

In Simple Terms

The problem it solves: Ethereum Layer 2 rollups need to post their transaction data back to Layer 1 for security. Before EIP-4844, they used calldata, which is permanently stored and expensive – making rollup transactions cost $0.10-$1.00+ each.
The blob solution: Proto-Danksharding introduces “blobs” – think of them as large, temporary data attachments. Blobs are much cheaper because they are automatically deleted after about 18 days and have their own separate fee market that does not compete with regular Ethereum transactions.
How rollups use it: Instead of cramming data into expensive calldata, rollups now attach their batched transaction data as blobs. The rollup smart contract on L1 receives a cryptographic proof (KZG commitment) that the data was available, without needing the actual blob data stored permanently.
The fee impact: After EIP-4844 went live in March 2024, Layer 2 fees dropped by 80-99%. Transactions on Arbitrum, Optimism, and Base that cost $0.10-$0.50 now cost fractions of a cent.
The path forward: Proto-Danksharding is “proto” because it is a stepping stone. Full Danksharding will use data availability sampling (DAS) to allow much more blob data per block without requiring every node to download all blobs, scaling Ethereum’s data capacity dramatically.

Real-World Examples

Scenario	Implementation	Outcome
Arbitrum fee reduction	Arbitrum switched from calldata to blob posting after Dencun upgrade	Transaction fees dropped from ~$0.25 to ~$0.001, making micro-transactions viable
Base network scaling	Coinbase’s Base L2 adopted blob transactions for data posting	Base became one of the cheapest L2s; daily transactions surged past 5 million
Optimism Superchain	Optimism and its Superchain (Base, Zora, Mode) all moved to blob data posting	Unified cost reduction across the entire Superchain ecosystem
KZG ceremony participation	Over 140,000 participants contributed randomness to the KZG trusted setup	Created one of the largest and most secure trusted setup ceremonies in cryptographic history

Advantages

Advantage	Description
Dramatic cost reduction	Layer 2 transaction fees dropped 80-99%, making rollups accessible to everyday users
Separate fee market	Blob gas pricing does not interfere with execution gas, preventing L2 data from driving up L1 fees
Temporary storage	Blob pruning after ~18 days prevents permanent state bloat on Ethereum nodes
Cryptographic foundation	KZG commitments provide the verification scheme needed for future full Danksharding
Rollup-agnostic	All rollup types (optimistic and ZK) benefit equally from cheaper data posting

Disadvantages & Risks

Risk	Description
Limited blob capacity	Initial 3-blob target (now 6 post-Pectra) may become insufficient as rollup demand grows
Data availability window	18-day pruning means data must be retrieved or archived before blobs expire
KZG trusted setup reliance	Blob commitments depend on the KZG ceremony; compromise of all participants would break security
Blob fee volatility	Blob fees can spike during high demand, temporarily increasing L2 costs
Not full solution	Proto-Danksharding is an intermediate step; full Danksharding with DAS is still needed for massive scale

Risk Management Tips:

Layer 2 users should monitor blob base fees during high-demand periods when costs may temporarily rise
Rollup operators should implement fallback to calldata if blob fees spike above calldata costs
Data indexers and archive nodes should ensure blob data is preserved before the pruning window
Monitor the transition from Proto-Danksharding to full Danksharding for upcoming capacity improvements
Diversify across multiple L2s to benefit from the lowest fees at any given time

FAQ

What is the difference between Proto-Danksharding and full Danksharding?

Proto-Danksharding (EIP-4844) introduces blobs with a limited capacity (3-9 per block). Full Danksharding will dramatically increase blob capacity (potentially to 64+ per block) using Data Availability Sampling (DAS), where nodes only need to download small random samples of blobs rather than full blobs, enabling massive data throughput.

Why are blobs deleted after 18 days?

Blobs are only needed temporarily for rollup security – the rollup uses the data to verify state transitions and allow fraud/validity proofs. After the challenge period passes, the actual data is no longer needed on Layer 1. Permanent storage would cause unsustainable node bloat.

What is a KZG commitment?

KZG (Kate-Zaverucha-Goldberg) is a polynomial commitment scheme that allows anyone to verify that blob data was correctly committed without downloading the full blob. It is computationally efficient and forms the cryptographic backbone of both Proto-Danksharding and future full Danksharding.

Did Proto-Danksharding affect Ethereum L1 gas fees?

Not directly. Proto-Danksharding created a separate fee market for blobs, so L1 execution gas fees remained independent. However, by moving rollup data from calldata to blobs, it freed up some calldata capacity, slightly reducing L1 fee pressure during peak times.

Can non-rollup applications use blobs?

Technically yes – any application can submit blob transactions. Some projects have experimented with using blobs for on-chain data storage, inscriptions, and other creative uses. However, blobs are primarily designed and priced for rollup data availability.