Ethereum's next major hard fork — codenamed Glamsterdam — is approaching a make-or-break moment. The Ethereum Foundation published its 2026 Protocol Priorities Update on February 18, formally organizing development into three tracks: Scale, Improve UX, and Harden the L1. The upgrade is targeted f...
"If you create a 10000 TPS EVM where its connection to L1 is mediated by a multisig bridge, then you are not scaling Ethereum." — Vitalik Buterin, Ethereum Co-founder
Ethereum's next major hard fork — codenamed Glamsterdam — is approaching a make-or-break moment. The Ethereum Foundation published its 2026 Protocol Priorities Update on February 18, formally organizing development into three tracks: Scale, Improve UX, and Harden the L1. The upgrade is targeted for May or June 2026 and carries the most ambitious technical scope since the Merge.
But ambition is colliding with reality. Developers acknowledged during a January All Core Developer Consensus call that two headliner features — enshrined Proposer-Builder Separation (ePBS, EIP-7732) and Block-level Access Lists (BALs, EIP-7928) — face serious testing and implementation challenges. If neither achieves cross-client interoperability by the end of February, both could be deferred to the Hegotá fork scheduled for the second half of 2026. The decision window is now measured in days.
This report examines the technical architecture of Glamsterdam, the economic implications of ePBS for the $3–7 billion annual MEV market, the validator economics at stake, and what a scope reduction would mean for Ethereum's competitive position against faster-moving Layer-1 rivals.
The February 18 protocol priorities update restructured Ethereum's development into three formal workstreams, each with dedicated leadership:
Scale — led by Ansgar Dietrichs, Marius van der Wijden, and Raúl Kripalani — is the engine room. It encompasses gas limit increases toward and beyond 100 million (up from 60 million after a 2025 raise), blob scaling for rollup data availability, and zkEVM attester client production readiness. The core Glamsterdam EIPs — EIP-7732 (ePBS) and EIP-7928 (BALs) — sit within this track.
Improve UX — led by Barnabé Monnot and Matt Garnett — targets native account abstraction, building on EIP-7702 (shipped in the May 2025 Pectra upgrade). The strategic goal is making every Ethereum wallet a smart contract wallet by default, eliminating reliance on external relayer infrastructure. Key proposals include EIP-7701 and EIP-8141 (Frame Transactions), which would embed smart account logic directly at the protocol layer.
Harden the L1 — led by Fredrik Svantes, Parithosh Jayanthi, and Thomas Thiery — is an entirely new track focused on preserving Ethereum's core properties as the network scales. It covers the "Trillion Dollar Security Initiative" for post-quantum cryptography readiness, FOCIL (EIP-7805) for fork-choice enforced inclusion lists to strengthen censorship resistance, and statelessness to reduce node hardware requirements.
The three-track structure is itself a signal. Ethereum's developers are conceding that scaling, usability, and security cannot be pursued as a single linear roadmap — they must be parallelized, with dedicated teams and independent timelines.
EIP-7732 is the crown jewel of Glamsterdam — and its most contentious component. Enshrined Proposer-Builder Separation would move the block-building marketplace from off-chain relay infrastructure directly into Ethereum's consensus layer.
The problem it solves is structural. Today, validators who want to maximize their block rewards use MEV-Boost, an off-chain sidecar built by Flashbots. MEV-Boost connects proposers (validators) with builders (entities that optimize transaction ordering for maximal extractable value) through trusted relays. The system works, but at a cost: two builders — Beaverbuild and Titan Builder — currently produce over 90% of Ethereum blocks. Flashbots' relay alone has historically facilitated roughly 84% of MEV-Boost blocks. This is extreme centralization in what is supposed to be a decentralized network.
The MEV market these entities service is enormous. Research estimates place annual MEV extraction on Ethereum between $3 billion and $7 billion, with sandwich attacks alone constituting $289.76 million — or 51.56% — of the $561.92 million in total MEV transaction volume recorded in 2025.
Under ePBS, the mechanics change fundamentally. Builders would assemble blocks and cryptographically seal their contents using a commit-reveal scheme. Proposers would select the highest-paying sealed block without being able to inspect or tamper with the transaction ordering inside. Transactions would only be revealed after the block is finalized, reducing the surface area for manipulation and censorship.
Critically, ePBS eliminates the trusted relay as an intermediary. The relay model requires both builders and proposers to trust a third party — relays must protect the builder's transaction bundle from being stolen by the proposer, while also ensuring the proposer receives the promised payment. This mutual distrust has made relay operation fragile and concentrated. By embedding PBS into the consensus layer, Ethereum would replace trust with cryptographic enforcement.
EIP-7928 is the less glamorous but equally consequential upgrade. Block-level Access Lists (BALs) would require blocks to pre-declare which accounts, storage slots, and smart contract data they will access during execution.
This seemingly administrative change unlocks two major capabilities. First, parallel execution: if the network knows in advance which state a block touches, non-overlapping transactions can be executed simultaneously rather than sequentially. This is how Solana and other newer chains achieve higher throughput, and BALs would bring Ethereum closer to parity without abandoning its security model.
Second, executionless state reconstruction: nodes could verify block state transitions without re-executing every transaction, dramatically reducing I/O overhead. This is foundational infrastructure for Ethereum's longer-term goal of supporting gas limits of 100 million and beyond — potentially reaching 200 million gas, which at the standard 21,000 gas per transfer would yield approximately 793 transactions per second.
For validators, BALs translate directly to lower hardware requirements and more predictable performance. Solo stakers and home validators — the constituency Ethereum most needs to preserve for credible decentralization — would benefit disproportionately.
Alongside the headliner EIPs, Glamsterdam includes a suite of gas repricing proposals designed to modernize Ethereum's fee structure:
Together, these proposals represent the most comprehensive gas model overhaul since EIP-1559. They carry real risk — smart contracts optimized for current gas pricing may break or become uneconomical — but they are necessary to make Ethereum's resource pricing reflect physical reality.
Here is where the narrative turns from aspiration to anxiety. During an All Core Developer Consensus call in late January 2026, developers confronted uncomfortable truths about implementation readiness:
The result is a potential "minimum viable Glamsterdam" — an upgrade stripped of its two most transformative features, shipping only the gas repricing suite and incremental improvements. This would be technically sound but strategically underwhelming. Ethereum would be upgrading its pricing model without upgrading its architecture.
Vitalik Buterin has acknowledged the pressure, noting that the original rollup-centric roadmap "no longer makes sense" and that Ethereum must scale directly on L1. He has also stated Ethereum can safely undergo up to four more major overhauls comparable to the Merge. But confidence in Ethereum's engineering capacity is not the same as confidence in Ethereum's engineering timeline.
If ePBS ships on schedule:
If ePBS is deferred:
The economic stakes are measurable. Validators running MEV-Boost currently earn approximately 60% more in staking rewards than those who don't. ePBS could redistribute these rewards more equitably, but deferral locks in the existing distribution hierarchy.
An academic analysis of ePBS's "free option problem" — where builders can decline to fill their committed slot during volatile moments — estimates option exercise at approximately 0.82% of blocks on average, rising to 6% during high-volatility days. This isn't trivial, but it's manageable and represents a known engineering challenge rather than a fundamental design flaw.
Glamsterdam crystallizes Ethereum's central tension: the network's engineering ambition consistently outpaces its engineering velocity. The three-track framework is a mature organizational response to this reality, but organization alone doesn't ship code.
The economic case for ePBS is overwhelming. A network that processes over $3 billion in annual MEV cannot indefinitely rely on two trusted builders and a handful of relays to intermediate that value. The commit-reveal mechanism in EIP-7732 is sound cryptographic engineering that would bring Ethereum's block market in line with its decentralization ethos.
But sound engineering doesn't matter if it isn't ready. The coming days will determine whether Ethereum ships its most consequential upgrade on schedule or defers its most important features to the back half of 2026. Either way, the subsidy-driven economic model identified in webthreepedia's foundational research persists: Ethereum generated approximately $3.1 billion in base-layer fees against $4–5 billion in annual staking inflation. Glamsterdam doesn't change that equation — but ePBS, if it ships, would at least ensure the value that does flow through the network is distributed more fairly.
The relay monopoly's days are numbered. The only question is whether they end in June or December.