Ethereum's Glamsterdam hard fork, tracked under meta-specification EIP-7773, bundles ten Ethereum Improvement Proposals into the network's largest protocol overhaul since the September 2022 Merge. The upgrade targets a 200 million gas limit per block — up from the current 60 million — parallel tr...
"This is the most significant protocol change since The Merge." — Parithosh Jayanthi, Ethereum Foundation Engineer
Ethereum's Glamsterdam hard fork, tracked under meta-specification EIP-7773, bundles ten Ethereum Improvement Proposals into the network's largest protocol overhaul since the September 2022 Merge. The upgrade targets a 200 million gas limit per block — up from the current 60 million — parallel transaction execution via Block-Level Access Lists, and the elimination of trusted relay infrastructure through Enshrined Proposer-Builder Separation. Devnet-7 completed stress testing in late July 2026. Mainnet activation is targeted for late August 2026, though September–December 2026 remains the firmer base case given cross-client testing requirements.
The economic implications are substantial. EIP-7904 recalibrates gas costs to reflect modern hardware, projecting a 78.6% reduction in L1 fees for complex smart contract interactions and up to 71% cheaper standard ETH transfers. EIP-7732's ePBS mechanism aims to reduce MEV-related losses by up to 70%, according to Flashbots Collective research. Ethereum's daily L1 fee revenue, which approached $30 million during 2021–2022, has declined to single-digit millions in 2026 — a compression that Glamsterdam's capacity expansion may or may not reverse, depending on whether demand scales with supply.
Glamsterdam ships ten proposals. The headliners are EIP-7732 (ePBS) and EIP-7928 (Block-Level Access Lists). The supporting cast addresses gas economics, contract limits, observability, and state management:
| EIP | Title | Function | |---|---|---| | EIP-7732 | Enshrined Proposer-Builder Separation | Moves block-building into the protocol layer | | EIP-7928 | Block-Level Access Lists | Enables parallel transaction execution | | EIP-7904 | Gas Cost Repricing | Realigns gas costs to computational complexity | | EIP-8037 | State-Creation Gas Cost Increase | Adds dedicated gas reservoir for state growth | | EIP-7778 | Block Gas Accounting Without Refunds | Removes refund-based gas accounting | | EIP-7708 | ETH Transfers Emit Logs | Adds event logs for native ETH movements | | EIP-7954 | Contract Size Increase | Raises max contract size from ~24 KiB to 32 KiB | | EIP-7843 | SLOTNUM Opcode | Exposes slot number to smart contracts | | EIP-8024 | Backward-Compatible SWAPN/DUPN/EXCHANGE | Extends EVM stack-manipulation opcodes | | EIP-7773 | Glamsterdam Meta Specification | Coordinates all fork-included proposals |
Two proposals were deferred: EIP-7782 (6-second slots) was removed from scope due to implementation complexity, and EIP-7805 (Fork-Choice Inclusion Lists) was moved to Hegotá, Ethereum's subsequent planned fork.
The current Ethereum block production pipeline relies on MEV-Boost, a sidecar relay system operated by Flashbots and a handful of other providers. Approximately 90% of Ethereum blocks pass through MEV-Boost relays as of mid-2026. This creates a trust dependency on relay operators and a centralization vector that the protocol does not directly govern.
EIP-7732 moves the proposer-builder handshake on-chain. Under ePBS, a proposer selects a builder's bid and publishes a signed commitment to the network. The builder then reveals the full block contents. The network validates that the builder's payload matches the commitment. No relay infrastructure is required.
The data propagation window expands from approximately 2 seconds to approximately 9 seconds, giving builders more time to assemble optimal payloads and reducing the advantage held by geographically co-located infrastructure. Flashbots Collective research estimates that ePBS could reduce MEV-related losses to users by up to 70%.
For validators, the change eliminates the requirement to run MEV-Boost relay software. For institutional staking operators like Figment, Lido, and Coinbase Cloud, this reduces the operational surface area — fewer third-party dependencies, fewer points of failure.
Ethereum currently executes transactions sequentially. EIP-7928 introduces Block-Level Access Lists (BALs) that record all accounts and storage locations accessed during block execution, along with their post-execution values. This metadata enables four capabilities: parallel disk reads, parallel transaction validation, parallel state root computation, and executionless state updates.
The parallelization opportunity is significant because 60–80% of transactions in a typical Ethereum block access disjoint storage slots. The remaining 20–40% can be processed using post-transaction state diffs rather than sequential execution.
The combined effect of BALs and the supporting state-management changes in EIP-8037 creates the technical foundation to raise the gas limit from 60 million to 200 million per block. At 200 million gas, Ethereum's theoretical throughput ceiling approaches 10,000 transactions per second — compared to roughly 15–30 TPS under current parameters.
EIP-8037 is the safety valve for this capacity expansion. It sets a fixed cost per byte of new state creation and carves out a separate gas reservoir for state growth, preventing the database from expanding past a manageable yearly budget even at higher gas limits.
Ethereum's current gas schedule assigns costs to opcodes based on benchmarks that, in some cases, date back to the network's 2015 launch. EIP-7904 revises the gas cost schedule for opcodes, precompiles, memory expansion, and data access to reflect actual computational complexity on modern hardware.
The recalibration produces measurable fee reductions:
Whether these fee reductions translate into lower user costs or higher network throughput at similar price points depends on demand elasticity. If the gas limit rises and demand fills the new capacity, the EIP-1559 base fee mechanism will adjust upward. The fee reduction is structural, not guaranteed.
The Glamsterdam development process has progressed through seven devnet iterations as of late July 2026:
After devnet completion, the fork must pass through two public testnets — Holesky and Hoodi — before mainnet activation. The Ethereum Foundation's internal working target is mainnet activation at the end of August 2026, but this date is aspirational. The firmer base case, accounting for the typical timeline between testnet forks and mainnet deployment, places activation in the September–December 2026 window.
Three risk factors could push activation into Q4 2026 or later: ePBS implementation complexity across multiple client teams, cross-client implementation parity (all execution and consensus clients must implement all ten EIPs identically), and the gas repricing rollout requiring extensive fuzzing.
Glamsterdam's relationship with Layer 2 rollups is indirect but consequential. The Dencun upgrade (March 2024) introduced blob transactions via EIP-4844, creating a separate data lane for rollups that cut L2 transaction costs by 90–99% — from $0.50–$5.00 to $0.001–$0.05.
Glamsterdam does not directly modify blob parameters. However, ePBS's extended payload-propagation window (from ~2 to ~9 seconds) creates structural room for future Blob Parameter Only (BPO) forks to increase blob capacity. More blobs mean lower data-posting costs for rollups — but also less fee pressure burning ETH on L1, contributing to the ongoing tension between Ethereum's L2 scaling strategy and its L1 fee-revenue model.
In early 2026, Vitalik Buterin questioned the pace of the rollup-only scaling approach, noting limited Layer 2 decentralization. The Ethereum Foundation subsequently reorganized its 2026 roadmap into three tracks — Scale, Improve UX, and Harden the L1 — with Glamsterdam as the first engineering expression of the pivot back toward high-performance L1 execution.
This pivot does not abandon L2s. It rebalances the architecture: a faster, cheaper L1 handles more activity natively while rollups continue to serve applications requiring even lower fees or specialized execution environments.
For institutional staking operations, Glamsterdam introduces two material changes.
First, ePBS eliminates the dependency on external MEV-Boost relays. Institutional operators currently run relay software alongside their validator clients — an additional operational component that introduces latency sensitivity and trust assumptions. Post-Glamsterdam, block-building competition occurs on-chain, reducing the infrastructure footprint for compliant validators.
Second, Glamsterdam includes changes to the validator exit queue mechanism. For large, concentrated exit requests — the kind institutional operators are more likely to encounter — processing that currently takes weeks could be completed in days post-upgrade. This improves capital efficiency for staking-as-a-service providers and ETF-linked staking products.
Figment's Q2 2026 Ethereum Validator Report noted that institutional stakers are running preparatory infrastructure updates for Glamsterdam compatibility, with most major operators targeting client upgrades in September.
Implementation risk: Ten simultaneous EIPs across four execution clients (Geth, Nethermind, Besu, Erigon) and five consensus clients (Prysm, Lighthouse, Teku, Nimbus, Lodestar) creates a combinatorial testing surface. The Merge shipped with fewer protocol changes.
Fee revenue uncertainty: Ethereum's daily L1 fees have compressed from near $30 million (2021–2022) to single-digit millions in 2026. A 3x capacity expansion via gas limit increases could further dilute fee revenue per unit of gas if demand does not scale proportionally. The impact on ETH burn rate under EIP-1559 is directionally negative unless utilization rises.
MEV redistribution: ePBS eliminates relays but does not eliminate MEV. Value extraction shifts from relay operators to an on-chain builder auction. Whether this reduces or merely redistributes centralization pressure among builders is an open empirical question.
State growth: Despite EIP-8037's state-growth gas reservoir, a 200M gas limit at sustained utilization could accelerate state bloat. Node operators with constrained hardware may face higher storage requirements.
Glamsterdam represents Ethereum's strategic pivot from an L2-only scaling narrative toward a dual approach: a higher-performance L1 paired with a rollup ecosystem. The technical scope — ePBS, parallel execution, gas repricing — is the most ambitious single-fork package since the Merge. Whether the capacity expansion reignites L1 fee demand or further compresses it is the central economic question. The answer depends on whether applications and users choose to move activity back to mainnet at lower costs, or whether rollups have already captured that demand permanently. The fork's September–December mainnet window leaves several months of public testnet data ahead. Until then, the upgrade remains a set of devnet-validated specifications, not production infrastructure.