Ethereum's next hard fork, Glamsterdam, targets mainnet activation in H1 2026 with two headline changes: Enshrined Proposer-Builder Separation (ePBS, EIP-7732) and Block-Level Access Lists (BALs, EIP-7928). The upgrade aims to dismantle an off-chain relay oligopoly where three block builders curr...
"2026 is the year that we take back lost ground in terms of self-sovereignty and trustlessness." — Vitalik Buterin, Co-Founder, Ethereum
Ethereum's next hard fork, Glamsterdam, targets mainnet activation in H1 2026 with two headline changes: Enshrined Proposer-Builder Separation (ePBS, EIP-7732) and Block-Level Access Lists (BALs, EIP-7928). The upgrade aims to dismantle an off-chain relay oligopoly where three block builders currently produce over 90% of Ethereum blocks, while simultaneously laying the foundation for parallel transaction execution and a gas limit increase from 60 million to 200 million per block.
If delivered on schedule, Glamsterdam would represent the most significant structural overhaul of Ethereum's execution and consensus layers since The Merge in September 2022. Developers on the All Core Devs calls have characterized the scope as targeting approximately 10,000 TPS on the base layer — roughly 10x current effective throughput — alongside a 78.6% reduction in gas costs through benchmarked repricing of opcodes. Devnet-4 testing of core EIPs is complete. The Ethereum Foundation DevOps team is transitioning to Devnet-5, with public testnets and dual audit phases planned for spring.
The risk profile is non-trivial. Base, the largest L2 by transaction volume, has publicly warned that bundling Fork-Choice Inclusion Lists (FOCIL) with ePBS could push the upgrade past 2026 entirely. Over 25 non-headliner EIPs remain under consideration, creating scope creep pressure. The tentative June 2026 target could realistically slip to Q3 or Q4.
Ethereum's block production has consolidated to a degree that contradicts the network's stated decentralization goals. According to data from Rated Network and CoinTelegraph, three builders — Beaverbuild, Titan Builder, and rsync-builder — controlled over 90% of block production through late 2024 and into 2025. In October 2024 alone, Beaverbuild and Titan Builder were responsible for 88.7% of all blocks.
The mechanism driving this concentration is access to private order flow. Approximately 60% of block value originates from private transaction streams unavailable to smaller builders. New entrants face a barrier-to-entry estimated at 1.4 ETH or more just to participate in block auctions, with the cost rising over time. The result is a winner-take-all dynamic where established builders attract more order flow because they win more auctions, and win more auctions because they have more order flow.
This dependency on off-chain relay infrastructure — primarily operated through Flashbots' MEV-Boost system — creates a trust assumption that Ethereum's consensus layer was not designed to accommodate. Flashbots historically relayed over 82% of all blocks. While the organization migrated operations to BuilderNet in December 2024, the underlying market structure remains concentrated.
The MEV extraction industry compounds the centralization concern. Between December 8, 2025 and January 6, 2026, MEV-extracting entities generated approximately $24 million in profit over 30 days on Ethereum. In 2025, sandwich attacks alone constituted $289.76 million — 51.56% of a total $561.92 million in MEV transaction volume. Since Ethereum's transition to proof-of-stake in September 2022, over 500,000 ETH (in excess of $1 billion) has been extracted via MEV strategies.
EIP-7732 is a consensus-layer change that formally separates the block proposer and block builder roles at the protocol level, eliminating the need for external relay infrastructure.
Under the current system, validators propose blocks, but the construction of those blocks — deciding which transactions to include and in what order — happens off-chain through third-party software. The proposer broadcasts a signed header; an external builder constructs the execution payload; and a relay mediates the exchange. This three-party handshake requires trust in the relay operator.
EIP-7732 replaces this with an in-protocol commit-reveal flow. The proposer commits to a block header. A separate builder constructs the execution payload. The protocol itself enforces builder commitments, introduces explicit deadlines, and handles builder non-delivery through fallback behavior — all without halting chain liveness.
The practical effect: any entity can participate as a builder without requiring a trusted relationship with a relay operator. Researchers estimate that ePBS could reduce MEV extraction by up to 70% by standardizing the handoff rules and creating a permissionless builder market. The relay layer, which currently represents a single point of potential censorship and failure, becomes architecturally unnecessary.
Vitalik Buterin has publicly outlined additional mechanisms intended to complement ePBS. FOCIL (Fork-Choice Enforced Inclusion Lists) would require 16 randomly selected attesters to each nominate a set of transactions that must appear somewhere in the block. A block that omits required transactions gets rejected. This mechanism is positioned for the subsequent Hegota upgrade rather than Glamsterdam itself, but its design is closely coupled with ePBS's architecture.
Buterin has also referenced encrypted mempools as a future direction — hiding transaction data until block finalization to eliminate frontrunning and sandwich attack windows entirely.
EIP-7928 introduces Block-Level Access Lists (BALs): structured declarations specifying which storage slots and accounts each transaction in a block will access. The purpose is to enable safe parallel execution of transactions by the Ethereum Virtual Machine.
Currently, Ethereum processes transactions sequentially within each block. Transaction N must complete before Transaction N+1 begins, because the EVM cannot know in advance whether two transactions touch the same state. This is the primary bottleneck limiting base-layer throughput.
BALs solve this by pre-declaring the accounts and smart contracts each block will interact with. When the network determines that two transactions access completely different parts of the state, it can process them simultaneously. Disk I/O, EVM execution, and post-state-root calculations become fully parallelizable, cutting worst-case block validation latency.
Combined with a gas limit increase from 60 million to 200 million per block, BALs target a throughput of approximately 10,000 TPS — up from the current effective rate of roughly 1,000 TPS on the base layer. The gas limit increase provides raw capacity; BALs provide the execution model that makes that capacity usable without proportionally increasing node hardware requirements.
This has direct implications for Ethereum's competitive position. Solana's theoretical throughput of 65,000 TPS (with real-world performance materially lower) has been a persistent comparison point. A 10x improvement on Ethereum's base layer narrows that gap while preserving Ethereum's validator decentralization — currently over 1 million active validators versus Solana's approximately 1,800.
EIP-7904 addresses a long-standing technical debt: Ethereum's gas costs for individual opcodes were calibrated years ago and no longer reflect the actual computational resources consumed on modern hardware.
The proposal adjusts gas costs for specific opcodes using empirical client benchmarks, targeting 18 operations that currently execute below 60 Mgas/s. By realigning gas prices with actual execution costs, the effective cost of using Ethereum drops without changing the gas limit.
The projected impact: a 78.6% reduction in gas fees for both simple ETH transfers and complex smart contract interactions. This compounds with already-low fees on the Ethereum mainnet. As of March 2026, Etherscan's gas tracker showed average prices around 0.196 Gwei, with lows touching 0.055 Gwei. ERC-20 transfers cost $0.01–$0.02, and Uniswap swaps approximately $0.14 — over 90% cheaper than a year ago. Glamsterdam's repricing would push these costs lower still.
The context matters: Ethereum's current low fees are substantially a consequence of activity migrating to L2s. L1 block usage sits at approximately 46%. Further fee reductions could either attract activity back to the base layer or, paradoxically, reduce fee revenue further for validators — a tension the Ethereum Foundation has not yet publicly addressed in the context of Glamsterdam.
The Ethereum Foundation's DevOps team has completed testing of three proposed EIPs on Devnet-4 and is transitioning to Devnet-5. Public testnets and dual audit phases are planned for spring 2026, with mainnet activation tentatively targeting June 2026.
Several risk factors threaten this timeline:
Scope creep. Over 25 non-headliner EIPs are under consideration for inclusion. Each additional EIP increases testing surface area and cross-client coordination burden. The Ethereum Foundation's Protocol Priorities Update, released February 18, 2026, organizes work around three pillars — Scale, Improve UX, and Harden the L1 — but the sheer number of proposals creates tension between ambition and deliverability.
FOCIL coupling. The Base engineering team — representing the largest Ethereum L2 by volume — has publicly warned that bundling FOCIL alongside ePBS could push the upgrade past 2026. FOCIL is now positioned for the Hegota upgrade (H2 2026), but its tight architectural coupling with ePBS means design decisions made in Glamsterdam constrain Hegota's options.
Cross-client complexity. ePBS is a consensus-layer change requiring coordinated implementation across all major Ethereum clients (Geth, Nethermind, Besu, Erigon on execution; Prysm, Lighthouse, Teku, Lodestar, Nimbus on consensus). Historical precedent suggests that cross-client coordination adds 2-4 months to optimistic timelines.
Validator risk. According to CryptoSlate's analysis, the 2026 roadmap includes a validator risk that developers characterize as "bigger than you think": the throughput gains from BALs and the gas limit increase require validators to process significantly more data per slot, potentially raising hardware requirements and squeezing smaller operators.
Viewed through an economic value lens, Glamsterdam's changes redistribute value flows across the Ethereum ecosystem in ways that merit close attention.
MEV redistribution. If ePBS reduces MEV extraction by the estimated 70%, the annual impact on MEV-extracting entities is material. Based on 2025 data ($561.92 million in total MEV volume), a 70% reduction implies roughly $393 million in annual extraction prevented — value that would instead remain with end users in the form of better execution prices.
Relay operator obsolescence. Flashbots, Blocknative, and other relay operators lose their structural role in the block production pipeline. BuilderNet's December 2024 launch pre-empted this shift, but the economic rationale for relay operation disappears under ePBS.
Validator economics. Lower gas costs and potential activity migration complicate the already-strained validator revenue model. Ethereum staking yields have compressed as the validator set expanded past 1 million. If Glamsterdam reduces fees while simultaneously making the base layer more competitive with L2s, the net effect on validator revenue depends on whether volume growth offsets per-transaction fee decline — a question without a clear answer at current data points.
L2 competitive dynamics. A 10x throughput increase on L1 challenges the value proposition of L2s built primarily on Ethereum's previous capacity constraints. L2s that differentiate on features (privacy, application-specific execution, governance) retain their positioning. Those that exist solely to offer cheaper execution face margin compression.
Glamsterdam is structurally the most ambitious Ethereum upgrade since The Merge. Unlike previous forks that focused on individual optimizations — Dencun's blob transactions, Fusaka's data availability improvements — Glamsterdam simultaneously rewrites block production (ePBS), execution (BALs), and pricing (gas repricing). The scope is comprehensive but risky.
The economic stakes are concrete. Block builder centralization, MEV extraction, and the relay oligopoly represent measurable costs imposed on Ethereum users. ePBS addresses these directly. The throughput and fee improvements address Ethereum's competitive position relative to alternative L1s and its own L2 ecosystem.
Whether the June 2026 timeline holds depends on whether developers maintain scope discipline. The Base team's FOCIL warning and the 25+ candidate EIPs suggest the pull toward feature expansion is substantial. Historical precedent favors delay over early delivery for Ethereum hard forks.
The data point that matters most will not come from testnets: it is whether block builder market share actually diversifies post-activation. If three entities still produce 90% of blocks after ePBS goes live, the protocol-level change will have failed its primary objective regardless of its technical elegance.