Ethereum's Glamsterdam hard fork — the network's most extensive protocol change since the September 2022 Merge — has entered final devnet testing with a mainnet activation window of Q3–Q4 2026. The upgrade bundles ten Ethereum Improvement Proposals (EIPs), headlined by Enshrined Proposer-Builder ...
"Probably the largest fork we've had since the Merge. It could change a lot of assumptions about Ethereum and set us up for much more scaling." — Parithosh Jayanthi, Ethereum Foundation Core Developer
Ethereum's Glamsterdam hard fork — the network's most extensive protocol change since the September 2022 Merge — has entered final devnet testing with a mainnet activation window of Q3–Q4 2026. The upgrade bundles ten Ethereum Improvement Proposals (EIPs), headlined by Enshrined Proposer-Builder Separation (EIP-7732) and Block-Level Access Lists (EIP-7928). Together, these changes restructure how blocks are built, executed, and verified on Ethereum's base layer.
The economic implications are material. The upgrade targets a gas limit increase from 60 million to 200 million per block, projects a ~78% reduction in L1 gas costs, and lays the groundwork for parallel transaction execution capable of approximately 10,000 TPS on the base layer — roughly 10x current throughput. For the $550 million annual MEV extraction market on Ethereum, the enshrinement of proposer-builder separation into the consensus layer eliminates the network's 90% dependency on third-party relay infrastructure such as MEV-Boost.
This report examines the technical architecture, economic consequences, and timeline risks of Glamsterdam, analyzing what the upgrade means for validators, L2 operators, builders, and end users in the context of Ethereum's broader value distribution.
Ethereum's block production currently depends on an extra-protocol system. Approximately 90% of mainnet blocks are constructed via MEV-Boost, where specialized builders assemble execution payloads and submit them through trusted relays to validators (proposers). The top three builders consistently produce over 50% of all blocks. The Flashbots relay alone processes more than 50% of PBS blocks.
EIP-7732 moves this handoff into Ethereum's consensus protocol. Under ePBS, block production splits into two formally defined roles:
The protocol locks in builder payments automatically, and a dedicated Payload Timeliness Committee verifies whether the builder disclosed the execution payload on time. The execution payload propagation window extends from approximately 2 seconds to 9 seconds, giving builders more room to construct optimized blocks while reducing missed-slot risk.
The result: the 80–90% dependency on third-party relays — Flashbots, BloXroute, Ultra Sound, and others — is eliminated at the protocol level. Relay trust assumptions (not stealing builder strategies, guaranteeing bid payment, revealing blinded blocks) become moot.
As Vitalik Buterin wrote on March 2, 2026: "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."
Ethereum currently processes transactions sequentially — one at a time, in order. EIP-7928 introduces Block-Level Access Lists, which function as a pre-execution map of every account, storage slot, balance, nonce, and code modification that a block will touch.
With BALs, nodes can identify which transactions have non-overlapping state dependencies before execution begins. Non-conflicting transactions can then be distributed across multiple CPU cores for parallel processing. The mechanism enables:
This changes Ethereum's execution model from a single-lane road to a multi-lane highway. Worst-case block validation latency drops substantially, enabling the network to handle higher throughput without proportionally increasing hardware requirements.
As of June 17, 2026, the Glamsterdam Meta-EIP (EIP-7773) lists ten proposals scheduled for inclusion:
| EIP | Description | |------|-------------| | EIP-7732 | Enshrined Proposer-Builder Separation (ePBS) | | EIP-7928 | Block-Level Access Lists (BALs) | | EIP-7708 | ETH transfers and burns emit logs | | EIP-7778 | Block gas accounting without refunds | | EIP-7843 | SLOTNUM opcode | | EIP-7954 | Max contract size increase from ~24 KiB to 32 KiB | | EIP-7976 | eth/70 partial block receipt lists | | EIP-8024 | Backward-compatible SWAPN, DUPN, and EXCHANGE opcodes | | EIP-8037 | State-creation gas cost increase | | EIP-8159 | eth/71 Block Access List Exchange |
Notable among the supporting EIPs: EIP-8037 raises the gas cost of state creation — a counterbalance to the overall gas limit increase that discourages state bloat. EIP-7954's contract size expansion from ~24 KiB to 32 KiB addresses a long-standing developer constraint. EIP-8024 introduces backward-compatible stack manipulation opcodes that improve EVM efficiency for compilers and smart contract optimization.
Glamsterdam targets a block gas limit increase from 60 million to 200 million — a 3.3x expansion of per-block capacity. Combined with BALs enabling parallel execution, the projected impact is a ~78% reduction in gas costs for both simple ETH transfers and complex smart contract interactions.
For context, current Ethereum L1 fees are already at historic lows. Average gas prices reached approximately 0.16 gwei by April 2026, with a basic ETH transfer costing under $0.01 and a Uniswap token swap running $0.03–$0.05. The low-fee environment is itself the product of prior upgrades — the Dencun upgrade delivered an approximate 95% fee reduction from prior peaks.
Glamsterdam pushes fees lower still, but the economic logic is volume-driven: cheaper execution is designed to attract more on-chain activity to L1, potentially increasing total fee revenue even as per-transaction costs decline.
Over $550 million in MEV is extracted annually from Ethereum alone. Under the current MEV-Boost system, value flows through a layered supply chain: searchers identify opportunities, builders assemble profit-maximizing blocks, relays intermediate, and validators capture a share via builder bids.
ePBS restructures this supply chain. By enshrining builder competition at the protocol level, it removes relay operators from the value chain entirely and creates a permissionless builder market. Researchers estimate ePBS could reduce total MEV extraction by up to 70%, translating to fairer execution pricing for traders, borrowers, and liquidity providers on L1.
However, Buterin himself acknowledged the limitation: ePBS "only prevents builder centralization from spilling over into the staking layer, but the issue of builder centralization itself remains." Solving builder-level centralization is deferred to future upgrades — specifically FOCIL (Forward Obligatory Commitment to Inclusion Lists) and encrypted mempools, slated for the Hegotá fork in late 2026 or early 2027.
The shift from relay-dependent PBS to protocol-native PBS carries structural implications for multiple stakeholders.
Relay operators — Flashbots, BloXroute, Ultra Sound Relay, Aestus, and others — face obsolescence of their core relay infrastructure. Flashbots has already begun this transition: in December 2024, it migrated all builders, orderflow, and refunds to BuilderNet and ceased operating centralized block builders on Ethereum. BuilderNet v1.2 launched in February 2025.
Validators/stakers gain more transparent MEV distribution. Currently, validators must trust relays to honor blinded block commitments and pay promised bids. Under ePBS, the protocol enforces these payments automatically. The removal of relay trust assumptions could accelerate staking participation among institutional operators that previously considered relay risk unacceptable.
Builders transition from a relay-gated market to a protocol-permissionless market. Entry barriers lower, but competition for block construction likely intensifies. The current concentration — where the top three builders produce >50% of blocks — may persist or deepen, as block construction efficiency advantages remain regardless of protocol-level changes.
Layer 2 networks settle data on Ethereum L1. When L1 block space becomes cheaper and more abundant, settlement costs for rollups decrease. This dynamic was demonstrated following the Fusaka upgrade in December 2025, which widened data capacity for rollups and compressed L2 fees.
Glamsterdam extends this pattern. Lower calldata costs and expanded block capacity reduce L2 settlement expenses, which flow through to end-user transaction fees on Arbitrum, Optimism, Base, and other rollup networks. L2 transaction fees on major networks already sit below $0.02; Glamsterdam could push them lower still.
The L2 ecosystem now accounts for approximately 95% of Ethereum's total transaction throughput, processing over 2 million daily transactions in 2026. The Ethereum ecosystem (L1 + L2s combined) has reached an all-time high TPS of 32,950. Cheaper L1 settlement broadens the economic viability of L2 use cases that were previously marginal — micropayments, high-frequency DeFi strategies, and gaming transactions among them.
Glamsterdam's development has followed a structured progression:
Remaining milestones:
The upgrade was originally targeted for H1 2026, with June as the aspirational window. After the Soldøgn Interop, Q3 2026 became the revised target. A slip into Q4 remains possible.
Builder centralization persists. ePBS eliminates relay trust but does not address the underlying concentration in block construction. The top builders maintain efficiency advantages from proprietary MEV search algorithms. Until FOCIL and encrypted mempools arrive in Hegotá, builder-level centralization remains unresolved.
Gas limit increase and state growth. A 3.3x gas limit expansion increases state growth rate. EIP-8037's state-creation gas cost increase partially offsets this, but long-term state management — including the still-delayed Verkle Trees migration — remains an open challenge. Verkle Trees are under candidate selection for Hegotá but have not passed security audits.
Parallel execution complexity. BALs introduce a new class of consensus-critical data. Bugs in access list generation or validation could cause chain splits or consensus failures. The multi-client devnet phase is designed to catch these, but the surface area is large.
Timeline risk. Ethereum's recent upgrade history shows consistent delays. Glamsterdam has already slipped from H1 to H2 2026. If public testnet issues emerge, mainnet could push into early 2027 — overlapping with Hegotá's own timeline.
Validator hardware requirements. Parallel execution benefits scale with CPU core count. Validators running minimal hardware may not capture the full performance gains, potentially widening the gap between institutional and independent validators.
Glamsterdam represents a structural reconfiguration of Ethereum's execution and consensus layers. The upgrade addresses two persistent inefficiencies: the protocol's reliance on off-chain relay infrastructure for block construction, and the sequential execution bottleneck that limits L1 throughput.
The economic stakes are quantifiable. A 78% gas fee reduction, 3.3x block capacity expansion, and elimination of relay trust assumptions reshape the cost structure for every participant — validators, builders, L2 operators, and end users. The $550 million annual MEV market faces its most significant protocol-level disruption since the introduction of MEV-Boost in 2022.
Whether Glamsterdam delivers on these projections depends on successful public testnet deployment and security review in the months ahead. The track record of Ethereum upgrades — functional but consistently delayed — suggests caution on timeline. The upgrade's technical ambition is not in question. Its execution timeline is.