As Ethereum prepares for its highly anticipated Dencun hard fork in early 2025, the developer community is zeroing in on a suite of critical Ethereum Improvement Proposals (EIPs) that promise to reshape scalability, security, and user experience. This upgrade marks a pivotal moment in Ethereum’s evolution—bringing tangible improvements that could redefine how users interact with Layer 2s, smart contracts, and decentralized applications.
With testnet forks scheduled across Goerli, Sepolia, and Holesky in early 2025, momentum is building toward mainnet deployment. While no major delays are expected, developers remain cautious, ensuring robustness before final activation.
Beyond Dencun, discussions have already begun on the next phase—tentatively dubbed Prague/Electra—where the community must decide between pursuing one large-scale core upgrade or bundling several smaller enhancements. A decision is expected post-Dencun, but for now, all eyes are on the transformative changes arriving this year.
Let’s explore the most impactful EIPs included in the Dencun upgrade and what they mean for Ethereum’s future.
EIP-4844: The Dawn of Proto-Danksharding
At the heart of Dencun lies EIP-4844, also known as Proto-Danksharding. This proposal is widely regarded as the most significant step toward solving Ethereum’s long-standing scalability bottleneck, especially for Layer 2 rollups.
👉 Discover how rollups are about to get 100x cheaper—click here to learn more.
Currently, rollups publish their transaction data directly onto Ethereum’s main chain (the Layer 1), which is secure but extremely costly. EIP-4844 introduces blobspace—a new type of data storage that allows rollups to attach temporary data blobs to blocks at a fraction of the current gas cost.
These blobs are stored only temporarily (around 18 days), reducing long-term state bloat while still maintaining data availability for verification. The result? Rollup transaction fees could drop by up to 90–100x, making decentralized apps more accessible than ever.
Lucas Henning, CTO at Suku, calls this “a breakthrough year for Web3.” With major L2s like Starknet, Optimism, and Arbitrum poised to adopt blob transactions immediately after activation, the impact will be felt across DeFi, gaming, and social platforms.
This isn’t full Danksharding yet—that’s years away—but EIP-4844 lays the essential groundwork by simulating its core mechanics.
Account Abstraction Gains Momentum: ERC-4337 & ERC-6900
While not part of the core protocol changes in Dencun, account abstraction (AA) is emerging as one of the most user-facing innovations shaping Ethereum’s future.
ERC-4337, launched in March 2023, introduced a standardized framework for AA without requiring a hard fork. It enables wallets to behave like smart contracts, unlocking features such as:
- Gasless transactions (sponsored by dApps)
- Social recovery (no more lost keys)
- Multisig security for everyday users
- Automated payments and recurring approvals
Now, ERC-6900 builds on this foundation by introducing delegated transactions. This standard allows users to grant limited authority to third parties—such as relayers or service providers—to execute specific actions on their behalf. Think of it as giving someone a signed blank check with strict limits.
The beauty of ERC-6900 is that it requires no consensus-layer changes, meaning adoption can happen rapidly across wallets and protocols. Developers gain flexibility; users gain control—all without compromising security.
This shift could finally bring crypto wallets into mainstream usability, bridging the gap between traditional finance UX and blockchain’s self-custody ethos.
EIP-1153: Efficient Temporary Storage for Smart Contracts
Smart contract developers often face a dilemma: storing data permanently on-chain is expensive and inefficient, especially when that data only needs to exist during a single transaction.
Enter EIP-1153, which introduces two new opcodes: TLOAD and TSTORE. These enable transient storage—a type of memory that persists only within the scope of one transaction and resets afterward.
Unlike traditional storage (which modifies Ethereum’s global state and incurs high gas fees), transient storage avoids state bloat entirely. This makes it ideal for use cases like:
- Flash loans
- Complex swap calculations
- Batched operations in DEXs
The Uniswap team has been a strong advocate for EIP-1153, planning to leverage it in Uniswap v4 to improve execution efficiency and reduce costs. By minimizing writes to permanent storage, v4 pools can become leaner, faster, and more gas-efficient.
This small but powerful change exemplifies how targeted optimizations can have outsized impacts on developer experience and network performance.
EIP-4788: Bridging Consensus and Execution Layers
Before EIP-4788, Ethereum’s execution layer (EVM) and consensus layer (Beacon Chain) operated largely in isolation. While validators secured the network via proof-of-stake, smart contracts had no direct way to verify consensus-level data—like validator sets or block roots—without relying on external oracles.
EIP-4788 solves this by committing the Beacon Chain’s block root into every EVM block header. In essence, it creates a trustless bridge between the two layers.
This means:
- Smart contracts can now verify consensus data natively
- No need for third-party oracle feeds
- Reduced attack surface and increased reliability
For liquid staking protocols like Lido, this is a game-changer. They can now build zkOracles that prove validator balances directly from consensus state—enabling more secure restaking solutions and cross-chain bridges.
Misha Komarov, founder of Nil Foundation, notes that “having access to consensus roots via EIP-4788 is absolutely helpful” for building zero-knowledge proofs that validate Beacon Chain activity inside the EVM.
👉 See how zk-powered oracles are transforming Ethereum security—click here to explore.
EIP-5656: Faster Memory Copying with MCOPY
Efficiency matters—even at the opcode level.
EIP-5656 introduces a new EVM instruction called MCOPY, designed to optimize how data is copied within memory during contract execution.
Currently, copying large chunks of data (like arrays or strings) requires looping through bytes or using inefficient patterns. MCOPY streamlines this process with a single, optimized operation—similar to memcpy in low-level programming languages.
Benefits include:
- Lower gas consumption for memory-heavy contracts
- Faster execution times
- Better support for complex dApps (e.g., AI-on-chain, video games)
Though less flashy than blob transactions or account abstraction, EIP-5656 represents vital plumbing work—improving Ethereum’s foundation so higher-level innovations can thrive.
EIP-6780: Taming the SELFDESTRUCT Opcode
The SELFDESTRUCT opcode has long been controversial. It allows a smart contract to delete itself from the blockchain and send its remaining ETH to a designated address.
While useful in theory, it poses real risks:
- Can be exploited for replay attacks
- Complicates state management
- Hinders future upgrades like stateless clients
EIP-6780 doesn’t remove SELFDESTRUCT entirely but limits its functionality. After activation, calling SELFDESTRUCT will no longer immediately wipe contract code or storage. Instead, it marks the contract for deletion during state cleanup cycles—a safer, more predictable approach.
This change supports Ethereum’s long-term vision of a leaner, more maintainable state—critical for achieving true scalability and enabling future upgrades like Verkle Trees.
Frequently Asked Questions (FAQ)
Q: What is the Dencun upgrade?
A: Dencun is a major Ethereum hard fork combining the Deneb (consensus layer) and Cancun (execution layer) upgrades. It focuses on improving scalability through proto-danksharding (EIP-4844) and enhancing developer tools via new EIPs.
Q: When will Dencun go live on mainnet?
A: Mainnet activation is expected in early 2025, following successful testnet rollouts on Goerli (Jan 17), Sepolia (Jan 30), and Holesky (Feb 7). No delays are currently anticipated.
Q: How will EIP-4844 affect Layer 2 users?
A: Users on rollups like Arbitrum, Optimism, and Starknet should see transaction fees drop by up to 90–100x due to cheaper data posting via blob transactions.
Q: Does account abstraction require a hard fork?
A: Not necessarily. Standards like ERC-4337 and ERC-6900 operate at the application layer and do not require protocol changes—making them faster to adopt.
Q: Is proto-danksharding the same as full sharding?
A: No. Proto-danksharding introduces blob-carrying transactions as a precursor to full sharding. True sharding—with 64 separate data shards—is still years away.
Q: Why limit SELFDESTRUCT instead of removing it?
A: Complete removal could break existing contracts. EIP-6780 strikes a balance by deprecating harmful behaviors while preserving backward compatibility during transition.
With Dencun on the horizon, Ethereum is entering a new era of efficiency, accessibility, and innovation. From slashing L2 costs to empowering smarter wallets and safer contracts, these EIPs collectively represent a leap forward—not just in technology, but in real-world usability.
As Prague/Electra planning begins, one thing is clear: Ethereum’s roadmap remains focused on sustainable growth, user empowerment, and long-term decentralization.