Ethereum's Glamsterdam hard fork, targeted for May-June 2026, represents the largest protocol-level overhaul since the September 2022 Merge. The upgrade centers on two headliner EIPs: EIP-7732, which enshrines proposer-builder separation directly into the consensus layer, and EIP-7928, which intr...
"In Glamsterdam, Ethereum is getting ePBS, which lets proposers outsource to a free permissionless market of block builders. This ensures that block builder centralization does not creep into staking centralization, but it leaves the question: what do we do about block builder centralization?" — Vitalik Buterin, Co-Founder, Ethereum
Ethereum's Glamsterdam hard fork, targeted for May-June 2026, represents the largest protocol-level overhaul since the September 2022 Merge. The upgrade centers on two headliner EIPs: EIP-7732, which enshrines proposer-builder separation directly into the consensus layer, and EIP-7928, which introduces block-level access lists enabling parallel transaction execution. Combined with EIP-7904's gas repricing — the first comprehensive opcode cost recalibration in years — the upgrade raises Ethereum's block gas limit from 60 million to 200 million and targets 10,000 transactions per second, roughly 10x current throughput.
The economic stakes are substantial. Ethereum generated approximately $2.73 billion in annual fee revenue as of early 2026, yet 80-90% of its block production currently depends on offchain builders, with two entities — Titan Builder (52%) and BuilderNet (28%) — constructing roughly 80% of all blocks. Glamsterdam attempts to pull this critical infrastructure back into the protocol, while simultaneously slashing gas costs by an estimated 78.6%. The implications for L2 economics, MEV extraction, and Ethereum's competitive position against alternative L1s are material.
Glamsterdam (a portmanteau of "Gloas" on the consensus side and "Amsterdam" on the execution side) is Ethereum's first hard fork following the Pectra (May 2025) and Fusaka (late 2025) upgrades. Vitalik Buterin outlined 8 Ethereum Improvement Proposals in late February 2026 that define the fork's scope.
The upgrade targets three structural problems simultaneously:
Four additional EIPs are under consideration: EIP-8037 (state creation gas cost increase), EIP-8038 (state-access gas cost increase), EIP-7954 (increased maximum contract size), and a set of gas repricing adjustments bundled under EIP-8007.
Under Ethereum's current architecture, block building happens offchain through a trust-based relay system. Proposers (validators) outsource block construction to specialized builders via MEV-Boost, a system developed by Flashbots. This works, but introduces dependencies on external infrastructure that sits outside the protocol's security model.
EIP-7732 moves the entire proposer-builder coordination onchain. Under ePBS, builders become staked protocol participants with onchain identity and signed bids. The proposer commits to a block header, and a separate builder constructs the execution payload. No external relay is required.
According to relay data from relayscan.io (February 2026), the relay market is currently split across relay.ultrasound.money (33.0% of payloads), titanrelay.xyz (21.1%), bloxroute.max-profit (19.4%), and bloxroute.regulated (12.4%). Flashbots' own relay holds just 3.5%. ePBS would eliminate the relay layer entirely.
Researchers estimate ePBS could reduce MEV extraction by up to 70%, translating into fairer execution pricing for DeFi users executing swaps, loans, and liquidity provision on Ethereum's base layer.
EIP-7928 introduces a mechanism for declaring all accounts and storage locations that will be accessed during a block's execution, along with their post-execution values. This seemingly administrative change enables four capabilities that are currently impossible:
BALs are constrained by the block gas limit, not a fixed item cap. With the gas limit rising from 60 million to 200 million, the capacity for parallel processing scales proportionally.
The practical effect: Ethereum can utilize modern multi-core hardware rather than processing transactions sequentially. Combined with ePBS, this creates a fundamentally different execution model. QuickNode's analysis of the upgrade notes that BALs make parallel execution "structurally viable" and enable future ZK-proof verification strategies.
EIP-7904 recalibrates gas costs for opcodes, precompiles, memory expansion, and data access based on empirical benchmarks from seven widely-used EVM implementations. The Gas Cost Estimator project, which measures actual computational effort per operation, provides the foundation.
Key changes include:
Many current gas prices were set during earlier Ethereum eras and no longer reflect execution costs on contemporary hardware. The repricing also introduces a structural separation: state creation gas is decoupled from execution gas, allowing compute to scale without proportionally increasing state bloat.
For developers, the practical implications include: financially incentivized leaner storage usage, fewer memory-related errors during compilation, and reduced smart contract security risks from gas-related edge cases.
The urgency behind EIP-7732 is quantifiable. According to builder statistics from rated.network, as of March 2026:
| Builder | Market Share | |---------|-------------| | Titan Builder | 52.3% | | BuilderNet (Flashbots) | 28.0% | | Quasar | 15.3% | | All others | 4.4% |
Three entities construct 95.6% of all Ethereum blocks. Titan Builder alone commands a majority. According to Cointelegraph reporting, Titan Builder has accumulated over $40 million in "hidden profit" from MEV-Boost, partly enabled by exclusive order flow agreements — notably with Banana Gun, which privately routes nearly all user transactions to Titan for block bundling.
Flashbots launched BuilderNet in November 2025 as a countermeasure, using a multi-operator system with Trusted Execution Environments to redistribute MEV back to users. BuilderNet has captured approximately 28% market share since launch, reducing but not eliminating the concentration problem.
On March 2, 2026, Buterin published a roadmap proposing Fork-Choice Inclusion Lists (FOCIL) as the next defense layer beyond ePBS. Under FOCIL, a randomly selected committee of validators creates a mandatory transaction inclusion list — blocks that ignore these transactions are rejected. Buterin proposed an expanded version, "Big FOCIL," that would handle nearly all transactions in a block, reducing the builder's role to state computation and minor MEV optimization.
Base, Coinbase's L2, has publicly opposed including FOCIL (EIP-7805) in Glamsterdam, citing risk that testing it alongside ePBS could push the upgrade past 2026, according to a blog post on blog.base.dev.
Glamsterdam's gas reduction and throughput increases have direct consequences for L2 economics. Base, Arbitrum, and Optimism together process nearly 90% of all L2 transactions, with Base alone surpassing 60% market share by late 2025. Base was the only L2 that turned a profit in 2025, earning approximately $55 million.
Lower blob posting costs (the fee L2s pay to settle data back to Ethereum mainnet) directly reduce operational expenses for rollups. Base's engineering team highlighted six non-headliner EIPs they support, including EIP-8070 (Sparse Blobpool), which reduces execution-layer node bandwidth consumption by approximately 4x through cell-level sampling.
The counter-dynamic: if Glamsterdam cuts L1 gas by 78% and MEV by up to 70%, some L2 activity could migrate back to Ethereum's base layer, which carries a higher security guarantee than any rollup. Pre-Dencun, average gas prices were approximately 72 gwei; post-Dencun, they fell to roughly 2.7 gwei. Glamsterdam targets roughly 1 gwei territory, making L1 competitive for transaction types that previously required L2s to be economically viable.
This creates a paradox: the upgrade simultaneously makes L2s cheaper to operate and L1 cheaper to use directly. Which effect dominates will determine whether the L2 oligopoly consolidates further or begins to contract.
As of March 2026, the Ethereum Foundation's DevOps team has completed testing of three proposed EIPs on Devnet-4. Development is transitioning to Devnet-5. Two dedicated testnets are active:
Public testnets with dual audit phases are planned for spring 2026. The technical scope is expected to be locked by late Q1 2026.
The target mainnet activation is May-June 2026, though developers have explicitly stated that hitting the date is secondary to correctness. Some delay into Q3 2026 remains possible. The subsequent upgrade, codenamed Hegota (Heze-Bogota), is planned for late 2026 and would include FOCIL if it does not ship with Glamsterdam.
Ethereum's fee revenue trajectory contextualizes what Glamsterdam means economically. The network generated approximately $2.73 billion in annual fees, but daily gas fees fell from a peak of $23 million to $6.3 million through 2025 as the Dencun upgrade shifted activity to cheaper L2s. Average gas prices dropped 93% year-over-year, from approximately 15.7 gwei to 1.1 gwei.
Glamsterdam's 78.6% further reduction in gas costs will compress L1 fee revenue per transaction even further. The thesis is that 10x throughput compensates: if transaction volume scales proportionally or faster, aggregate fee revenue stabilizes or grows despite lower per-transaction costs. This mirrors the Dencun experience, where lower fees attracted sufficient new activity to partially offset the price decline.
The MEV question is equally consequential. Sandwich attack extraction fell from nearly $10 million monthly in late 2024 to approximately $2.5 million monthly by October 2025, with average per-attack profit declining to just $3. Only six attackers generated more than $10,000 in total profit throughout 2025. ePBS aims to further constrain this already-declining extraction channel, but whether it eliminates or merely redistributes MEV value remains an open research question.
ETH traded at approximately $2,073 on March 26, 2026 — down roughly 56% from its cycle highs — creating a disconnect between protocol-level engineering progress and token market performance. Whether Glamsterdam's throughput and cost improvements close that gap depends on whether the upgrade drives measurable new demand for Ethereum blockspace.
Glamsterdam is an engineering upgrade with economic consequences. It addresses the structural fact that Ethereum's block production has consolidated into three entities controlling 95.6% of all blocks — a centralization level that contradicts the network's foundational design. Simultaneously, it restructures how gas is priced, how transactions execute, and how much throughput the network can absorb.
The economic question is whether a 10x capacity increase at 78% lower cost generates sufficient new demand to sustain or grow Ethereum's $2.73 billion annual fee base. History from Dencun suggests lower costs attract more activity, but the magnitude of Glamsterdam's changes — parallel execution, enshrined PBS, comprehensive gas repricing — introduces variables without precedent in Ethereum's upgrade history.
The upgrade does not resolve all open questions. FOCIL, the mechanism Buterin proposed to commoditize block building entirely, is deferred. The L1-vs-L2 migration dynamics remain uncertain. And ETH's 56% drawdown from cycle highs suggests the market is not yet pricing Glamsterdam as a catalyst.
What Glamsterdam does is raise the technical floor. If Ethereum's capacity increases 10x and its costs fall 78%, the failure mode shifts from "the network cannot scale" to "the network scaled but demand did not follow." That is a different — and arguably better — problem to have.