Ethereum is preparing to reorganize the data structure that underpins its network. The plan is EIP-8297, known as the Partitioned Binary Tree, or PBT. The proposal aims to enable smaller state proofs and a path to state expiry without taking the network offline.
The proposal recently passed a meaningful test. On October 29, 2026, PBT went live on a mainnet shadowfork called glamsterdam-pbt-1, run by ethPandaOps.
What the Partitioned Binary Tree changes
Ethereum’s “state” is the full record of every account balance, every smart contract’s code, and every piece of data those contracts store. Today, Ethereum organizes that record using a hexary Merkle Patricia Trie, or MPT.
EIP-8297 swaps that for a binary tree. Every branch splits in two, which makes the structure simpler to walk through and simpler to prove things about. The proposal also folds three separate categories into one structure: account headers, contract storage, and contract code would all live in a single binary tree.
The “partitioned” part refers to how keys get sorted. They are grouped into designated zones, which keeps related data closer together. The design also drops RLP encoding, the serialization format Ethereum currently uses. Removing it addresses some of the proving headaches the existing setup creates.
The shadowfork test and timeline
A shadowfork copies the real mainnet and runs it separately, letting developers rehearse an upgrade against live conditions without putting actual funds at risk. On glamsterdam-pbt-1, the switch to PBT went smoothly. Nodes running Besu, the client built by Consensys, rebuilt their state into the new format directly from their existing databases. That reconstruction took approximately 2 hours and 41 minutes.
Full mainnet deployment of PBT is slated for Q4 2026, and no specific activation date has been announced.
How we got here
EIP-8297 was created on June 11, 2026. It reflects a collaborative effort that includes contributions from Ethereum co-founder Vitalik Buterin. The proposal builds on earlier research into binary trees and Verkle trees, both explored as possible replacements for the current trie.
One of its central design choices is invisibility. The upgrade is not meant to change gas schedules or alter how the EVM operates. Existing contracts and users should not notice anything different. Wallets keep working, contracts keep running, and nobody has to rewrite code.
What this means for Ethereum
The most direct beneficiaries are the people building toward stateless clients—nodes that can verify blocks without storing the full state themselves. Stateless clients depend on compact proofs. If PBT delivers smaller state proofs and simpler validity proofs as designed, it removes one of the bigger obstacles in that roadmap.
State expiry is the other big piece. Ethereum’s state keeps growing, and every full node has to carry it. PBT is built to support future state expiry methods while keeping existing access patterns intact.
For the zero-knowledge crowd, eliminating RLP encoding is a quiet win. Proving systems have struggled with the current format, and a cleaner structure should make that work less painful. Since PBT is explicitly designed not to change gas schedules, any cost benefits would likely arrive through later upgrades it enables rather than on activation day.
The risks are mostly about execution. Only Besu’s results from the shadowfork are detailed here, and Ethereum runs on multiple client implementations that all need to convert their state correctly.
What to watch next: additional shadowfork rounds, results from other execution clients, and an announced mainnet activation date. Until that date lands, Q4 2026 remains a target rather than a commitment.
