As 2026 gets underway, Ethereum has clearly transitioned from a phase of experimental development to one of rigorous, industrial-scale hardening. After the successful deployment of the Pectra and the Fusaka upgrades in 2025, the network has settled into a predictable bi-annual release cadence, akin to a “train schedule” where upgrades launch reliably every six months. This operational maturity is a big deal. It gives the core protocol room to stay stable, while allowing infrastructure providers and Layer-2 (L2) ecosystems to align their growth strategies with the protocol’s standardized roadmap.
Last year was defined by “The Surge”, specifically unlocking massive data availability for rollups to lower transaction costs. In 2026, the focus shifts. The attention is turning toward “The Scourge” and Layer-1 (L1) performance. The ecosystem is no longer just focused on expanding capacity for L2s; it is now turning inward to address the centralization risks that emerged following The Merge. The reliance on trusted relays for block production and the bottlenecks of sequential execution have become the primary targets for this next generation of protocol improvements.
Addressing all these challenges head-on, the Glamsterdam upgrade marks the first major milestone of this new chapter for Ethereum. It is not merely an incremental tweak, but a structural re-engineering of how the network processes transactions and secures its consensus.
By confronting the long-standing reliance on trusted middleware and breaking free from the Ethereum Virtual Machine’s (EVM’s) single-threaded constraints, Glamsterdam pushes Ethereum toward stronger censorship resistance and true parallel execution. In doing so, it lays the groundwork for a genuinely trustless global settlement layer. One that is built not just to scale, but to endure.
What is Ethereum Glamsterdam upgrade
Announced at the ACDE (All Core Devs Execution) 226, the Glamsterdam upgrade represents Ethereum’s first major hard fork of 2026, tentatively scheduled for the first half of the year. The name itself reflects the two sides of the protocol it touches: “Gloas,” which is named after a star, and the execution layer upgrade, “Amsterdam,” following Ethereum’s tradition of Devconnect cities.
Glamsterdam marks a strategic pivot in Ethereum’s development roadmap. After a 2025 largely defined by scaling data capacity for L2s (the “Surge”), the attention is now turning back to L1. The primary goals are to enhance execution efficiency and enshrine censorship resistance directly into the protocol.
Operating under Ethereum’s new bi-annual “train schedule,” Glamsterdam is designed to deliver specific, high-impact architectural changes. It moves the network away from relying on trusted third parties for block production and prepares the system for a future of parallel processing.
Key developments within the upgrade
The Glamsterdam upgrade is built around two major “headliner” features that fundamentally reshape how Ethereum processes transactions and builds blocks.
Enshrined Proposer-Builder Separation (ePBS – EIP-7732) Currently, Ethereum validators rely on third-party relays (like MEV-Boost) to source profitable blocks from builders. While effective, this creates a centralization risk and a trusted choke point. ePBS integrates this market directly into the consensus protocol, allowing validators (Proposers) and Builders to transact trustlessly.
The mechanism splits the 12-second slot into two phases. First, during the consensus phase, the Proposer commits to a block header without seeing the transactions, preventing them from stealing the Builder’s MEV strategies. Then, in the reveal phase, the Builder discloses the full payload. A newly introduced “Payload Timeliness Committee” (PTC) ensures the Builder reveals the data on time, slashing them if they fail.
Block-Level Access Lists (BALs – EIP-7928) For most of Ethereum’s history, the EVM has processed transactions sequentially, one after another. This limitation exists because the network cannot predict which state (account balances or storage) a transaction will modify until it actually runs. BALs solve this by mandating that every block includes a comprehensive list of all accounts and storage slots accessed.
By declaring these “access paths” upfront, the network can identify which transactions are independent of each other. This allows the node software to verify non-conflicting transactions simultaneously on multiple CPU cores (parallel execution) and pre-fetch data from disk in parallel, significantly boosting Layer 1 throughput.
Significance, Advancing and Futureproofing
What makes Glamsterdam truly significant is how directly it defends Ethereum’s core value: neutrality. By removing the need for trusted relays via ePBS, the protocol reduces the power of any single entity to censor transactions or coerce validators. It ensures that the market for block building remains open and permissionless.
On the technical side, Glamsterdam tackles one of Ethereum’s longest-standing limitations: the “sequential bottleneck.” The introduction of BALs acts as the plumbing for a Parallel EVM. This efficiency gain is projected to allow the block gas limit to rise safely from 60 million to 80-100 million, lowering costs for users and Layer 2s alike.
These changes are critical prerequisites for future-proofing the next phase of the roadmap. BALs structure data in a way that simplifies the creation of cryptographic proofs. This is a necessary step toward Statelessness, where nodes can verify the chain without storing terabytes of state data.
Overview on past upgrades and their link to Glamsterdam
To understand Glamsterdam, one must look at the foundation laid by the upgrades of 2025.
Pectra (May 2025) focused on optimizing the actors within the system. It introduced EIP-7251, which raised the maximum effective balance for validators to 2,048 ETH. This consolidation of validators reduced peer-to-peer (P2P) networking noise, creating the bandwidth “headroom” necessary for subsequent complex upgrades.
Fusaka (December 2025) focused on optimizing capacity. It implemented PeerDAS (EIP-7594), a sampling technology that allowed the network to massively scale data availability for L2 rollups.
The link is direct: Fusaka solved the data problem, ensuring rollups have enough space to grow. Glamsterdam now solves the execution problem, ensuring L1 can compute fast enough to manage that growth, while securing the trust layer of block production.
The road ahead for Ethereum
After Glamsterdam, Ethereum’s focus now shifts to the Hegota upgrade (Heze + Bogota), scheduled for the second half of 2026. Hegota is expected to implement Verkle Trees, completing Ethereum’s transition to stateless clients.
Additionally, Hegota may also pick up features that didn’t quite make the cut for Glamsterdam. One of the most notable is FOCIL (Fork-Choice Enforced Inclusion Lists), which is designed to force builders to include censored transactions, serving as the final line of defense for the network’s neutrality.
More broadly, this upgrade reflects how much Ethereum has fully adopted a predictable release cadence. Features that aren’t ready for a specific “train” like Glamsterdam are simply pushed to the next one, ensuring the protocol continues to evolve steadily without the long delays that characterized its early history.
