Skip to content

EIP-7623

EIP-7623

Overview

EIP-7623 (Increase Calldata Cost) is part of the Ethereum Pectra upgrade, proposing an increase in calldata costs for data-intensive transactions to reduce the maximum possible block size and control block size variance. This proposal was activated on the Ethereum mainnet in May 2025.

The core objective of this proposal is to limit the use of Ethereum block space by data-intensive applications such as rollups, while not affecting the transaction costs of ordinary users.

Background and Motivation

Block Size Problem Under the current calldata pricing: - Cost per byte of calldata: 4 gas (zero bytes) / 16 gas (non-zero bytes) - Block Gas Limit: 30,000,000 gas - Theoretical maximum block size: approximately 7.15 MB - Actual average block size: approximately 100 KB

The Enormous Gap - Maximum possible: 7.15 MB - Actual usage: 100 KB - Gap: over 70x

This enormous gap leads to: 1. Large block size variance: Blocks can be enormous at peak times, affecting network stability 2. Resource waste: Nodes must prepare resources for extreme cases 3. Potential DOS attacks: Malicious actors can construct very large blocks

Rollup Data Demand Since EIP-4844 introduced blobs, some rollups still use calldata: - Some rollups have not yet migrated to blobs - Calldata provides faster data availability - But it occupies primary block space

EIP-7623 Solution

Differentiated Pricing Increases calldata costs for data-intensive transactions: - Data-intensive transactions: Calldata cost raised to 10/40 gas/byte - Computation-intensive transactions: Maintains the original 4/16 gas/byte

Classification Criteria How to distinguish between the two types of transactions: 

If a transaction's EVM computation gas < a certain ratio of calldata gas
-> Classified as a data-intensive transaction
-> Uses higher calldata pricing

Pricing Mechanism

Old Pricing (Before EIP-7623) 

Zero bytes: 4 gas/byte
Non-zero bytes: 16 gas/byte

Maximum calldata: 30,000,000 / 4 ≈ 7.15 MB

New Pricing (Data-Intensive Transactions) 

Zero bytes: 10 gas/byte
Non-zero bytes: 40 gas/byte

Maximum calldata: 30,000,000 / 40 = 0.75 MB

Computation-Intensive Transactions Remain Unchanged If a transaction has sufficient EVM computation: 

Still uses 4/16 gas/byte
Ordinary users are unaffected

Technical Implementation

Gas Calculation Formula 

Standard calldata gas = zero byte count x 4 + non-zero byte count x 16

If (total gas - calldata gas) < calldata gas x coefficient:
    // Data-intensive transaction
    Actual gas = zero byte count x 10 + non-zero byte count x 40
else:
    // Computation-intensive transaction
    Actual gas = standard calldata gas

Threshold Design The proposal sets a threshold to ensure: - Transactions with sufficient EVM computation are unaffected - Pure data transactions pay higher costs - Balances the needs of both types of users

Impact Analysis

Impact on Rollups

Rollups Using Calldata - Data publishing costs increase significantly (2.5x) - Strong incentive to migrate to blobs (EIP-4844) - May increase transaction fees for L2 users

Rollups Already Migrated to Blobs - Completely unaffected - Blobs use a separate fee market - Continue to enjoy low-cost data availability

Impact on Ordinary Users

*DeFi* Transactions** - Involve significant EVM computation (swaps, liquidity operations, etc.) - Computation gas >> calldata gas - Completely unaffected, maintains 4/16 gas/byte

Simple Transfers - ETH transfer: 21,000 gas, no calldata - ERC-20 transfer: ~50,000 gas, minimal calldata - Impact is negligible

Contract Deployment - Deployment involves a large amount of bytecode - But also includes creation and initialization computation - Most cases are unaffected

Data-Intensive Applications - Pure data storage or proof submissions - Costs increase 2.5x - Should consider using blobs or other solutions

Impact on Block Size

Maximum Block Size 

Old scheme: 30M / 4 = 7.5 MB
New scheme: 30M / 40 = 0.75 MB

Reduction: approximately 90%

Practical Impact - Maximum block size in extreme cases is significantly reduced - Block size variance is lowered - Network stability is improved

Node Requirements - Reduced peak bandwidth requirements - Slower storage growth rate (in conjunction with EIP-4444) - Improved sustainability of node operations

Relationship with Other EIPs

EIP-4844 (Blob Transactions) - Provides rollups with a low-cost data availability alternative - EIP-7623 incentivizes rollups to migrate from calldata to blobs - Both work together to achieve scaling

EIP-4444 (Historical Data Expiry) - Reduces the volume of new calldata - Works alongside historical data pruning - Controls long-term storage growth

EIP-7691 (Blob Count Increase) - Provides more blob space - Rollups have sufficient blob capacity - Makes migration from calldata more viable

Implementation Status

Activation Information - Upgrade name: Pectra (Prague-Electra) - Activation time: May 7, 2025 - Activation epoch: 364032 - Network: Ethereum mainnet

Client Support All major clients have implemented this: - Geth - Nethermind - Besu - Erigon

Community Response

Reasons for Support - Reduces uncertainty in maximum block size - Incentivizes more efficient data availability usage - Does not affect ordinary users - Improves network stability

Concerns Raised - May impact rollups that have not yet migrated to blobs - Increases costs for certain applications - Complex pricing mechanism

Final Consensus After thorough discussion, the community concluded that the benefits outweigh the costs: - Blobs already provide a better alternative - Limiting block size variance benefits decentralization - Ordinary users are unaffected

Migration Recommendations

For Rollups 1. Assess current status - Check whether calldata is used to publish data - Calculate the cost impact of EIP-7623

  1. Migrate to Blobs
  2. Implement blob transaction support
  3. Test blob data availability

  4. Gradually switch to blobs

  5. Optimize data

  6. Compress calldata
  7. Optimize batch processing
  8. Reduce unnecessary data

For Developers 1. Check applications - Assess whether they are data-intensive - Calculate cost impact

  1. Consider alternatives
  2. Use off-chain storage + proofs
  3. Use blobs to store large amounts of data

  4. Optimize data structures

  5. Test and monitor

  6. Validate on testnets
  7. Monitor cost changes
  8. Adjust gas estimation

Long-Term Impact

Data Availability Market - Blobs become the primary data availability solution - Calldata is used for small amounts of data requiring immediate availability - Clear functional separation

Network Health - More predictable block sizes - Lower node operating costs - Greater degree of decentralization

*Rollup* Ecosystem** - Accelerates migration to blobs - Lower L2 operating costs (via blobs) - More sustainable scaling solution

Economic Model

Fee Burning Calldata Gas fees burn ETH: - Data-intensive transactions burn more - But total transaction volume may decrease - Net impact depends on rollup migration speed

*Gas* Market** - Calldata and blob fee markets are separated - Rollups can choose the most economical solution - Market-driven resource allocation

Monitoring Metrics

Key Metrics - Average block size - Maximum block size - Calldata usage - Blob adoption rate - Node sync time

Expected Results - Maximum block size significantly reduced - Blob usage rate increases - Network becomes more stable