Ethereum’s Pectra Upgrade May Be Split Into Two Forks

With Ethereum’s forthcoming Pectra upgrade shaping up to be the most expansive in the network’s history, developers are considering splitting its deployment over two separate hard forks.

Following the latest All Core Devs Execution call on June 6, Christine Kim, a researcher at Galaxy Digital, tweeted that developers are exploring executing the upgrade over two forks.

Kim noted that nine Ethereum Improvement Proposals (EIPs) are already confirmed for inclusion, with an additional five upgrades under consideration. She added that the upgrades are being staged in batches across devnet implementations, enabling Pectra’s scope to expand in the lead-up to deployment.

“It is A LOT for one upgrade, so there's still a good chance Pectra gets split,” Kim said. “Staging multiple EIPs for implementation on devnets is a new development for the Ethereum governance process that leaves the door open for an upgrade with an ever-changing scope.”

Pectra’s upcoming Devnet 1 will include the same eight EIPs tested on Devnet 0.

Disruptive changes

According to Vitalik Buterin, Ethereum’s chief scientist, Pectra is among the last upgrades expected to introduce disruptive changes to Ethereum.

“After scaling is done, we’re in a stage where the hardest and fastest parts of the Ethereum protocol’s transition are essentially over,” Buterin said during an Ethereum Foundation livestream in April 2023. “Various things will need to be done, but those… things can be safely done at a slower pace.”

Five execution layer upgrades and four consensus layer upgrades have been “finalized” for inclusion as part of Pectra so far, spanning an eclectic assortment of EIPs dating as far back as 2020.

Verkle Trees and statelessness

EIP-2935 is a critical component of Pectra required to facilitate the introduction of Verkle Trees and statelessness. Verkle Trees will remove the need for nodes to store the network’s state locally, significantly reducing the computational requirements on validators.

According to Kim, developers are also working on three related “parallel upgrades” that are likely to be bundled into Pectra.

These include the development of validator light clients, which do not download the entire Ethereum blockchain. Light clients are intended to improve the decentralization of Ethereum by allowing users to validate the network using "resource-constrained devices such as tablets and cell phones.”

With nodes no longer required to store the entire block history of Ethereum, EIP-4444 would also formalize the deletion of historic data from full nodes after a set amount of time, further reducing the computational demands for validators in a bid to improve node decentralization.

“With EIP-4444: History Expiry, you would not have to store most of the history locally,” said Buterin. “The amount of data that you would need to be a node would decrease from multiple terabytes to... being able to run a node in RAM.”

DApps and other nodes that need to access deleted block history would be able to query third-party data services following EIP-4444’s deployment.

Developers are also exploring PeerDAS, which would provide further Layer 2 data availability following the introduction of proto-danksharding in March.

Consensus layer upgrades

Many of the upgrades set for inclusion in Pectra seek to improve the functionality and efficiency of Ethereum staking.

EIP-6110 aims to decrease the delay between stakers depositing collateral on the Ethereum mainnet execution layer and the corresponding transaction processing on the Beacon Chain. It would also reduce the complexity of validator client software.

EIP-7251 aims to slow the growth and reduce the size of Ethereum’s validator set to bolster the network's security. EIP-7002 would expand the number of designs available to developers creating staking pools.

EIP-7549 targets improving the efficiency of attestations, which would reduce network load and node bandwidth demands.

Execution layer upgrades

On the execution layer, EIP-2537 and EIP-7865 would bolster interoperability between the Ethereum mainnet and its Beacon Chain consensus layer. This would expand the functionality of decentralized staking pools, DAOs, and dApps that communicate with the Beacon Chain for staking functions.

A bundle of ten EIPs relating to EVM Object Format (EOF) aims to facelift the Ethereum Virtual Machine, Ethereum’s core smart contract engine. EOF would improve the upgradability of EVM contracts alongside the efficiency of smart contract code execution.

EIP-7702 is a controversial upgrade that would enable “smart account” features for regular externally owned account (EOA) wallets. The upgrade would enable batched transaction execution, transaction fee sponsoring, and delegated wallet security, but critics have warned the upgrade could make users vulnerable to new attack vectors.

“It's important to note that devs are staging the implementation of Pectra EIPs, meaning if EOF or EIP-7702 or PeerDAS don't make it into a devnet over the coming months, devs may drop it entirely from the upgrade or split out Pectra into two hard forks,” Kim said.

Unconfirmed upgrades

Kim noted two additional consensus layer upgrades that developers have not yet confirmed for inclusion in Pectra.

EIP-7688 would make it easier for staking pools to verify whether a participating validator has been slashed.

Developers may also consider reducing the rate of new Ether issuance. Mike Neuder, an Ethereum Foundation researcher, recently advocated for such a move, arguing that the Ethereum network and ecosystem has undergone “Seismic” change since Ether’s staking rewards mechanism was finalized amid the lead-up to the Beacon Chain’s deployment in 2020.

However, the proposal quickly proved divisive, with many community members pushing back against Nueder’s suggestion. Some onlookers feared it could open the door to Ethereum Foundation researchers frequently meddling with the network’s tokenomics without input from the broader community.