Ethereum's Glamsterdam hard fork — the network's largest protocol change since The Merge in September 2022 — is nearing final scope lock after 100+ core contributors completed the Soldøgn interop in Svalbard, Norway on May 2, 2026. Three deliverables emerged from that event: a post-upgrade gas li...
"2026 is the year that we take back lost ground in terms of self-sovereignty and trustlessness." — Vitalik Buterin, Ethereum Co-Founder
Ethereum's Glamsterdam hard fork — the network's largest protocol change since The Merge in September 2022 — is nearing final scope lock after 100+ core contributors completed the Soldøgn interop in Svalbard, Norway on May 2, 2026. Three deliverables emerged from that event: a post-upgrade gas limit floor of 200 million (roughly 3.3x the current 60 million), stable multi-client ePBS implementations with external builders operational, and finalized EIP-8037 state-creation repricing parameters.
The original target of H1 2026 mainnet activation has slipped. Developers now point to Q3 2026 as the realistic window, with public testnet deployments and dual audit phases still ahead. The upgrade's two headliner EIPs — EIP-7732 (Enshrined Proposer-Builder Separation) and EIP-7928 (Block-Level Access Lists) — aim to restructure how Ethereum builds blocks and executes transactions, with projected impacts of up to 10,000 TPS throughput and a 78.6% reduction in gas fees.
ETH trades at approximately $1,648 with a market capitalization of $199 billion as of June 11, 2026. The token has underperformed Bitcoin year-to-date as fee revenue flowing to the base layer has declined more than 95% from 2021 highs, a dynamic Glamsterdam partially addresses by enabling substantially higher L1 throughput.
The Soldøgn interop — held the week of April 28 to May 2, 2026, in Longyearbyen, Svalbard (78°N) — served as the final scope-definition event. Teams initially targeted a 4 EL × 4 CL devnet by Monday but achieved a stable 4×3 configuration by Tuesday for stress testing. By Friday, nearly all clients were running on glamsterdam-devnet-2 with end-to-end external builder pipelines operational.
A separate devnet track for Block-Level Access Lists ran in parallel, with specifications stabilizing on bal-devnet-6. EIP-8037 repricing parameters, initially following a dynamic model, were shifted to a fixed cost_per_state_byte approach early in the week and stabilized at the opcode level by Thursday.
The Ethereum Foundation's DevOps team has since completed Devnet-4 testing and transitioned to Devnet-5. Three of the proposed EIPs have been validated on Devnet-4. A new test suite revealed an execution-request invalidation gap across all client implementations — a finding that underscores the complexity of shipping ePBS.
Timeline projection: The original H1 2026 target remains aspirational. Multiple developer teams now expect a July-September window, contingent on public testnet validation. The Ethereum Foundation has confirmed devnets are running stably but has not committed to a specific mainnet date.
EIP-7732 is the consensus-layer centerpiece of Glamsterdam. It introduces a protocol-native separation between block proposers and block builders, replacing the external MEV-Boost relay system that currently mediates over 92% of Ethereum blocks.
How it works: The EIP decouples execution validation from consensus validation, both logically and temporally. A new in-protocol entity — the builder — commits to producing a valid block by submitting a payment hash to the proposer. Validators receive a new duty: submitting payload timeliness attestations to verify builder performance.
What it replaces: Since 2022, Ethereum has relied on off-protocol relays — principally MEV-Boost operated by Flashbots and others — to coordinate block building. As of late October 2025, relay market share was distributed: Ultra Sound relay at 32.3%, Titan Relay at 24.75%, bloXroute relays combined at approximately 26%, and Flashbots' own relay at just 3.44%. Approximately 94% of MEV-Boost blocks were dominated by two builders: Beaverbuild and Titan.
Why it matters: ePBS establishes protocol-native blockspace auctions and cryptoeconomic accountability. It guarantees that an honest proposer receives payment regardless of builder behavior, and that an honest builder's payload becomes the canonical chain head regardless of proposer action. This removes the trust assumption currently embedded in relay infrastructure.
The design is explicitly modular. According to the EIP specification, ePBS is "designed not as a standalone change, but as a modular foundation that can support and interoperate with other major upgrades in Ethereum's roadmap."
EIP-7928 is the execution-layer counterpart. It introduces Block-Level Access Lists (BALs) — records of all accounts and storage locations accessed during block execution, along with their post-execution values.
The parallelization thesis: Ethereum currently executes transactions sequentially because the EVM cannot know in advance which addresses and storage slots a transaction will touch. BALs solve this by explicitly listing the state accessed or modified by each transaction.
According to research published on the Ethereum Magicians forum and ethresear.ch, 60-80% of Ethereum transactions access disjoint storage slots, meaning they can be parallelized immediately. The remaining 20-40% can still be parallelized using post-transaction state diffs contained in the BAL.
Technical overhead: A new field is added to the block header: block_access_list_hash, containing the Keccak-256 hash of the RLP-encoded access list. Historical analysis shows an average BAL size of approximately 70 KiB. BAL verification occurs alongside parallel IO and EVM operations without delaying block processing.
Throughput implications: BALs enable parallel disk reads, parallel transaction validation, parallel state root computation, and executionless state updates. Combined with the 200 million gas limit floor, developers project a path toward 10,000 TPS on L1 — compared to approximately 15-30 TPS today.
EIP-8037 addresses a structural risk: if the gas limit triples to 200 million without adjusting state-creation costs, Ethereum's state size could grow unboundedly, threatening node operability.
The proposal implements a fixed cost_per_state_byte model that targets roughly 60 GiB of state growth per year at 300 million gas. New account creation and state-heavy contract deployments become 8-10x more expensive, while standard transaction and call operations remain unaffected.
The repricing shifted from a dynamic to a fixed model during the Soldøgn interop. Accounting moved from mid-execution to end-of-call-frame on Monday, underwent iterative refinements through Wednesday, and stabilized at the opcode level by Thursday. Parameters are now finalized and deployed on bal-devnet-6.
This EIP represents a direct intervention in the economics of state growth — the kind of resource-pricing discipline that separates sustainable protocol economics from subsidy-dependent architectures.
Glamsterdam's economic significance extends beyond gas fee reduction. Several dynamics converge:
L1 fee revenue recovery: Ethereum's L1 fee revenue has collapsed since EIP-4844 (Dencun, March 2024) slashed L2 data costs by over 90%. In the first year after Dencun, blob transactions paid approximately 1,020 ETH in blob base fees plus roughly 2,000 ETH in EIP-1559 fees — a total of about $8 million. Current gas prices sit at 0.04-0.05 Gwei, compared to 1.67 Gwei a year prior. A 3.3x increase in the gas limit floor, combined with higher throughput demand, could partially reverse this trend by expanding the total fee-addressable transaction volume.
L2 dynamic: Layer 2 networks now account for approximately 95% of Ethereum's total transaction throughput, processing about 2 million daily transactions. L2s currently pay minimal blob fees to L1. EIP-7918, expected in a subsequent upgrade, introduces a reserve price for blob fees to establish a sustainable floor.
MEV redistribution: According to 2026 data, over $550 million is extracted annually from Ethereum alone. Monthly toxic extraction is down roughly 75% from late 2024 levels as private mempools and front-running protection tools gained adoption. ePBS brings MEV auctions on-chain, potentially capturing a portion of this value at the protocol level rather than directing it through off-protocol relay infrastructure.
The transition from off-chain relays to enshrined PBS represents one of the most consequential infrastructure changes in Ethereum's history.
The current architecture routes approximately 92% of Ethereum blocks through MEV-Boost, with 94% of those blocks built by just two entities. Flashbots acknowledged this centralization risk by launching BuilderNet in November 2024 — a decentralized block-builder running in trusted execution environments, co-developed with Beaverbuild and Nethermind. In December 2024, Flashbots migrated all builders, orderflow, and refunds to BuilderNet and ceased operating centralized block builders.
ePBS makes this entire relay layer optional. Block construction becomes a protocol-level function with cryptoeconomic guarantees rather than a trust-based intermediary service. The question for the MEV supply chain is whether builders, searchers, and orderflow providers will transition smoothly to the enshrined mechanism or whether parallel systems persist during a migration period.
Several features originally considered for Glamsterdam were reassigned to Hegotá, a follow-on upgrade now positioned as a late-2026 or early-2027 cleanup fork:
Leadership transitions are also underway. Alex Stokes, a key contributor, began a three-month sabbatical after Soldøgn. Pari assumed interim ACDC moderation, Barnabas took over ACDT, and Nixo and Ansgar became ACDE co-chairs.
Glamsterdam is an infrastructure upgrade, not a narrative event. It restructures two fundamental protocol functions — block construction and transaction execution — that have operated essentially unchanged since the Beacon Chain launch. The 200 million gas floor, parallel execution via BALs, and enshrined PBS represent the most aggressive L1 scaling package Ethereum has attempted.
The economic question is whether these changes arrive in time. Ethereum's L1 fee revenue has cratered post-Dencun, L2s capture the majority of transaction activity, and ETH has underperformed BTC year-to-date. Glamsterdam expands L1 capacity but does not, by itself, guarantee demand will fill it.
The MEV restructuring carries the clearest near-term impact. Moving from a two-builder oligopoly intermediated by off-protocol relays to an enshrined protocol mechanism changes the trust model for every participant in Ethereum's block production pipeline. Whether that transition is orderly or disruptive will depend on execution quality across client teams — the same teams that just discovered an invalidation gap during interop testing in the Arctic.