Definição
The Dencun Upgrade (pronounced “Den-kun”) is a landmark Ethereum network upgrade that activated on the Ethereum mainnet on March 13, 2024, at epoch 269,568, combining changes to both the Ethereum execution layer (Cancun) and the consensus layer (Deneb) – hence the portmanteau “Dencun.” The upgrade’s most significant and widely impactful change is the implementation of EIP-4844 (Proto-Danksharding), which introduced a new transaction type called “blob-carrying transactions” that allows Ethereum Layer 2 rollups (Optimism, rede Arbitrum, Base, zkSync, Starknet, etc.) to post their transaction data to Ethereum in cheaper, temporary “blob” storage rather than expensive calldata. This single change reduced Layer 2 transaction fees by 10x to 100x almost overnight – with transaction costs on networks like Base dropping from cents to fractions of a cent. The “blob” data is a 128KB temporary storage type that is available for approximately 18 days before being pruned, sufficient for Camada 2 dispute-resolution windows without requiring permanent archival by all Ethereum nodes. Beyond EIP-4844, Dencun included additional improvements to Ethereum’s execution and consensus layers: EIP-1153 (transient storage opcodes for gas-efficient within-transaction state), EIP-4788 (beacon block root in EVM for cross-layer access), EIP-5656 (MCOPY opcode for efficient memory operations), and EIP-7044/7045/7514/7516 improvements to staking and validator operations. Dencun represents Ethereum’s most impactful usability upgrade since the Merge.
Leia também: zkSyncName
Origem e história
| Data | Evento |
| fevereiro 2022 | EIP-4844 “proto-danksharding” formally proposed by Vitalik Buterin, Dankrad Feist, and others |
| 2023 | EIP-4844 becomes centerpiece of Cancun upgrade planning; extensive testing on devnets |
| Jan 2024 | Goerli testnet Dencun upgrade activates (January 17) |
| Jan 2024 | Sepolia testnet upgrade activates (January 30) |
| fevereiro 2024 | Holesky testnet upgrade activates (February 7) |
| 13 de março de 2024 | Dencun activates on Ethereum mainnet at 13:55 UTC |
| 13 a 14 de março de 2024 | First blob transactions processed; L2 fees drop 10–100x immediately |
| Pós-março de 2024 | Base processes millions of daily transactions at sub-cent fees; Coinbase attributes massive Base growth to Dencun |
| 2024-2026 | Blob market matures; EIP-7691 (blob count increase) added to Pectra upgrade for further capacity |
“Dencun is the moment Ethereum’s L2 ecosystem became genuinely affordable for everyday users. Blob fees transformed what’s possible.”
Como Funciona
| EIP | Nome | Propósito | Impacto |
| EIP-4844 | Proto-Danksharding | Blob transactions for rollup DA | 10–100x L2 fee reduction |
| EIP-1153 | Armazenamento transitório | TSTORE/TLOAD opcodes | Gas-efficient within-tx storage |
| EIP-4788 | Beacon Block Root | Expose consensus layer root to EVM | Cross-layer smart contract access |
| EIP-5656 | MCOPY opcode | Efficient memory-to-memory copy | Gas optimization for contracts |
| EIP-6780 | SELFDESTRUCT limits | Restrict selfdestruct to same-tx | Safer contract lifecycle management |
| EIP-7044 | Perpetual signed exits | Validator exit message validity | Staking UX improvement |
| EIP-7514 | MAX_CHURN_LIMIT | Limit validator queue growth | Staking stability |
| EIP-7516 | BLOBBASEFEE opcode | Read blob gas price in EVM | Rollup contract optimization |
Em termos simples
- Cheaper L2 transactions: Before Dencun, sending a token on Arbitrum cost ~$0.50–$2. After Dencun, the same transaction costs $0.001–$0.05. This makes Ethereum’s Layer 2 ecosystem practical for everyday micro-transactions.
- Blobs as cheap parking: Instead of storing rollup data in expensive permanent calldata, blobs provide a temporary “parking space” for rollup data that’s automatically deleted after ~18 days – cheap because it doesn’t need to be stored forever.
- KZG commitments for verification: Even after blob data is deleted, the KZG commitment (a mathematical fingerprint) remains permanently on-chain. This allows anyone to verify that data was available at the time, satisfying Layer 2 security requirements.
- Proto-danksharding as a step: EIP-4844 is “proto” danksharding because it implements the blob transaction type but not the full peer-to-peer sampling and massive throughput of complete danksharding. It’s the first step in Ethereum’s long-term data availability scaling roadmap.
- Impacto no mundo real: Base (Coinbase’s L2) processed 80+ million transactions per month after Dencun at fees that compete with Solana – demonstrating that Ethereum L2s can serve mass-market use cases.
Ver mais: Resumo otimista: Confie primeiro, verifique apenas se necessário
Exemplos do mundo real
| Cenário | Implementação | Resultado |
| Base transaction fees | Post-Dencun blob pricing on Base | Average fee drops from ~$0.50 to ~$0.002 – 250x reduction |
| Adoção de blobs do Arbitrum | Arbitrum switches calldata → blobs immediately after Dencun | User savings estimated at $50M+ monthly compared to pre-Dencun costs |
| Coinbase retail adoption | Base’s ultra-low fees enable social and micro-payment use cases | Farcaster, DEGEN token, and social crypto boom on Base |
| zkSync blob posting | zkSync Era posts compressed state diffs as blobs | ZK rollup costs become competitive with optimistic rollups |
| Starknet optimization | Starknet uses blobs for state diff posting | L3 and advanced ZK applications become economically viable |
Vantagens
| A Vantagem | Descrição |
| Dramatic fee reduction | 10–100x reduction in L2 transaction fees immediately upon activation |
| L2 ecosystem growth | Lower costs enable mass-market applications previously unviable on L2s |
| Ethereum DA scaling | First step toward Ethereum becoming a scalable data availability layer |
| Temporary storage efficiency | 18-day data retention matches L2 dispute windows without permanent archival burden |
| KZG security | Cryptographic commitments maintain security guarantees without permanent blob storage |
Desvantagens e Riscos
| Desvantagem | Descrição |
| Temporary data retention | Blobs deleted after ~18 days – external archival services needed for long-term data access |
| Blob market congestion | During high-demand periods, blob fees can spike (multiple rollups competing for limited blob space) |
| Proto-danksharding limitations | Initial blob capacity (3–6 per block) is still limited; full danksharding needed for maximum throughput |
| Complexity increase | Blob transaction type adds new node requirements and complexity to Ethereum’s architecture |
Dicas de gerenciamento de risco:
- For L2 applications requiring historical data access beyond 18 days, ensure archival strategy using services like EthStorage, Portal Network, or third-party indexers
- Monitor blob fee market during high-activity periods – extreme congestion can temporarily reduce L2 fee benefits
- Understand that Dencun is the first step; subsequent upgrades (Pectra with EIP-7691, eventually full danksharding) will further expand capacity
Perguntas frequentes
What is the difference between calldata and blob data?
Calldata is the traditional way rollups post data to Ethereum – stored in transaction inputs permanently across all full nodes, priced per byte at a premium. Blobs are a new data type introduced by EIP-4844: larger (128KB per blob), temporary (~18 days), stored in a separate blob pool rather than the permanent chain state, and priced by a separate fee market – making them 10–100x cheaper per byte for rollup purposes.
How many blobs can each Ethereum block contain after Dencun?
Dencun initially supports a target of 3 blobs per block with a maximum of 6. Each blob is 128KB, meaning up to 768KB of blob data per block. This is a significant increase from pre-Dencun capacity, though the subsequent Pectra upgrade (EIP-7691) increased blob counts further.
Does Dencun make Ethereum itself cheaper for regular users?
Dencun primarily benefits Layer 2 users – not direct Ethereum mainnet users. Mainnet transaction fees (gas) are determined by calldata and execution costs, which Dencun didn’t change directly. However, by making L2s much cheaper, Dencun indirectly encourages migration to L2s where fees are dramatically lower.
What is KZG polynomial commitment and why does Dencun use it?
KZG commitments are a cryptographic primitive that creates a compact mathematical “fingerprint” of a large data blob. This fingerprint is stored permanently on-chain even after the blob is pruned. Anyone with the necessary commitment can verify claims about the original data’s content and availability, providing a fraud-proof foundation for rollup security without requiring permanent blob storage.
What comes after Dencun in Ethereum’s roadmap?
The Pectra upgrade (2024–2026) follows Dencun, with EIP-7691 increasing blob count targets. Full danksharding – featuring 2D DAS with potentially 128+ blobs per block – is the longer-term goal, potentially multiplying current blob capacity by 20x or more. Additional upgrades include Verkle trees and statelessness for further Ethereum efficiency improvements.
