Ethereum's next hard fork, Glamsterdam, will raise the network's gas limit from 60 million to 200 million, introduce enshrined proposer-builder separation (ePBS) via EIP-7732, and enable parallel transaction processing through block-level access lists (EIP-7928). The combined effect: a projected ...
"We don't friggin need more copypasta EVM chains." — Vitalik Buterin, Ethereum co-founder, February 2026
Ethereum's next hard fork, Glamsterdam, will raise the network's gas limit from 60 million to 200 million, introduce enshrined proposer-builder separation (ePBS) via EIP-7732, and enable parallel transaction processing through block-level access lists (EIP-7928). The combined effect: a projected 3.3x increase in execution capacity, up to 78% lower gas fees, and the elimination of Ethereum's 90% dependency on off-chain relay infrastructure for block construction.
Originally targeted for June 2026, the upgrade has slipped toward Q3 after the Soldøgn interop devnet in Svalbard, Norway, concluded on May 2 with stable multi-client implementations but unresolved edge cases. Public testnet activations on Holesky and Sepolia have not yet been scheduled. If Glamsterdam delivers on its specifications, it represents the largest single capacity expansion since The Merge in September 2022.
Ethereum's mainnet currently operates at a 60 million gas limit per block, a level validators signaled for in November 2025 after more than 513,000 validators voted in favor. Average gas prices have fallen to 0.05–0.12 gwei as of May 2026, with average transaction fees at $0.16–$0.22, according to Etherscan and ycharts data. For context, average fees sat near $1.85 in mid-2025.
The decline reflects a structural shift. Layer 2 networks now handle approximately 95% of Ethereum's total transaction throughput, per The Block's data from late 2025. Mainnet has become a settlement and data availability layer rather than a primary execution venue for retail users.
The block construction market is heavily concentrated. Approximately 92% of Ethereum blocks are built via MEV-Boost, the off-chain middleware that outsources block assembly to specialized builders. Within that system, two builders — Beaverbuild and Titan — assemble roughly 94% of all MEV-Boost blocks, according to Rated Network data. This creates a trust dependency on a small number of off-chain relays: Ultra Sound relay handles 32.3% of MEV-Boost payloads, Titan Relay 24.75%, bloXroute relays combined approximately 26%, and Flashbots' own relay just 3.44%.
Glamsterdam simultaneously updates the Execution Layer (Amsterdam) and the Consensus Layer (Gloas). Three headline EIPs anchor the release:
| EIP | Component | Function | |-----|-----------|----------| | EIP-7732 | Consensus Layer | Enshrined proposer-builder separation (ePBS) | | EIP-7928 | Execution Layer | Block-level access lists for parallel processing | | EIP-7904 | Execution Layer | Gas repricing aligned to actual computational costs |
Over 25 additional non-headliner EIPs are under consideration. EIP-8037 — a state creation gas cost model — has been finalized and locked in during the Soldøgn interop.
EIP-7732 moves the proposer-builder handoff from off-chain relay infrastructure directly into the Ethereum consensus protocol. Under the current architecture, validators rely on external software (MEV-Boost) and trusted relays to receive block payloads from builders. This creates a dependency on intermediaries that operate outside protocol rules.
Under ePBS, builders submit bids directly to the consensus layer. The protocol manages the block-building auction within each slot through a commit-reveal mechanism. Builders become staked protocol participants with on-chain identity. Validators join a new Payload Timeliness Committee (PTC) to verify that builders reveal block contents within slot deadlines.
The practical effect: the 80–90% reliance on third-party relays becomes optional rather than structural. The existing MEV-Boost relay infrastructure continues to function — there is no forced migration — but the new in-protocol builder market provides a trust-minimized alternative.
For validators, operational complexity decreases. Under MEV-Boost, validators must manage relay connections, evaluate builder bids through off-chain software, and make rapid decisions within tight slot timing windows. Under ePBS, the validator role reduces to: propose the block, attest to the chain, participate in consensus.
For builders, the competitive landscape opens. More builders can compete directly within protocol rules rather than negotiating access through a small number of relay operators.
Block-level access lists (BALs) address what developers identify as Ethereum's primary throughput bottleneck: state access latency. Currently, Ethereum clients only discover which accounts and storage slots a block touches during execution. The EVM can complete computation quickly, but execution stalls on sequential disk reads as the client fetches state data from a database that grows larger every day.
EIP-7928 requires every block to include a manifest of all accounts and storage slots that its transactions will access, along with post-execution values. With these explicit mappings, three operations that currently run sequentially — disk I/O, EVM execution, and post-state root calculation — become parallelizable.
According to developer Toni Wahrstätter, who authored the EIP, the proposal received broad support from core devs during recent All Core Devs calls. The BAL specification stabilized on bal-devnet-6 during the Soldøgn interop, where developers iterated through multiple gas accounting approaches before settling on opcode-level accounting.
The throughput target: 10,000 transactions per second on Layer 1, roughly 10x Ethereum's current capacity.
EIP-7904 realigns gas costs with actual computational resource consumption. Many current gas prices were calibrated years ago and no longer reflect execution costs on modern hardware. The recalibration produces an estimated 78% reduction in gas costs for both simple ETH transfers and complex smart contract interactions. According to MEXC research, a Uniswap trade that currently costs $3–$8 in gas could fall below $1 post-Glamsterdam.
EIP-8037 introduces a fixed cost_per_state_byte model targeting state database growth of approximately 60 GiB per year at a 300 million gas limit. Contract deployment costs rise roughly 10x and new account creation approximately 8.5x. The proposal was initially designed with dynamic per-state-byte pricing tied to the block gas limit, but during the Soldøgn interop, teams agreed to drop dynamic pricing in favor of a fixed cost, with future repricing handled at fork boundaries.
The combined 200 million gas limit floor — agreed upon at Soldøgn — represents a 3.3x increase from the current 60 million ceiling. Ethereum core contributors achieved this target through convergence of ePBS optimizations, BAL improvements, and EIP-8037 repricing adjustments.
| Milestone | Date | Status | |-----------|------|--------| | Soldøgn Interop (Svalbard) | May 2, 2026 | Completed | | Multi-client devnet (glamsterdam-devnet-2) | May 2026 | Running | | EIP-8037 repricing numbers | May 2026 | Locked | | 200M gas limit consensus | May 2026 | Achieved | | Holesky testnet activation | TBD | Not scheduled | | Sepolia testnet activation | TBD | Not scheduled | | Mainnet activation | Q3 2026 (est.) | Target |
The Soldøgn interop gathered over 100 Ethereum core contributors in Longyearbyen, Svalbard, compressing approximately one month of asynchronous development into each day. By the end of the week, nearly all clients were running together on glamsterdam-devnet-2 with the external builders pipeline tested end-to-end.
The Ethereum Foundation announced concurrent leadership changes on May 11, 2026. Tim Beiko and Barnabé Monnot are departing the Foundation, while Alex Stokes will take a sabbatical. Will Corcoran, Kev Wedderburn, and Fredrik were appointed as new Protocol Cluster leads. Corcoran will oversee zkVM proofs and post-quantum consensus coordination; Wedderburn leads zkEVM development; Fredrik manages the "Trillion Dollar Security" initiative.
The FOCIL (Fork-Choice Inclusion Lists) controversy has been resolved by deferring FOCIL to the subsequent Hegotá upgrade, which is targeted for late 2026 and will also introduce Verkle Trees to reduce node storage requirements by up to 90%.
For Layer 2 rollups, Glamsterdam expands blob capacity — the dedicated data space rollups use to post compressed transaction data to Ethereum. More blob capacity reduces bidding wars for data availability and lowers Layer 2 settlement costs. With the 200 million gas limit and expanded blobs, rollups could potentially handle hundreds of thousands of transactions per second while maintaining Ethereum's security guarantees.
For institutional stakers, the transition is material. According to Figment's analysis, ePBS simplifies the validator technology stack by removing the need to manage off-chain relay software. Institutional staking operations that currently maintain MEV-Boost infrastructure can eventually transition to a protocol-native system. However, validators will need to update client software and prepare for PTC duties.
For protocol economics, gas repricing creates a tension. Lower fees per transaction should increase transaction volume, but the net effect on validator revenue depends on whether volume growth offsets per-unit fee compression. With average fees already at $0.16–$0.22, further reductions could pressure validator economics absent substantial demand increases.
Scope risk. Over 25 non-headliner EIPs remain under consideration. Each addition increases testing surface area. The Base team's earlier warning about FOCIL scope creep was addressed by deferral, but the number of remaining candidate EIPs suggests the risk persists.
ePBS implementation complexity. The Ethereum Foundation's May 2026 protocol update noted that ePBS implementation "is proving to be trickier than anticipated." Multi-client interoperability remains an active area of debugging.
Builder adoption. ePBS is opt-in. If the existing relay infrastructure continues to offer competitive execution quality, builders may have limited incentive to migrate to the in-protocol system. The intended decentralization benefits depend on meaningful builder participation.
State growth trade-offs. EIP-8037's 8.5–10x cost increase for account creation and contract deployment could deter certain categories of applications, particularly those that create large numbers of accounts or deploy contracts frequently.
Leadership transition. The simultaneous departure of Beiko, Monnot, and Stokes from active Foundation roles during a critical pre-fork period introduces coordination risk, though the new leads were involved in the Soldøgn interop.
Glamsterdam is an execution-layer overhaul packaged as an upgrade. It addresses three structural constraints simultaneously: block construction centralization, sequential state access bottlenecks, and misaligned gas pricing. The 200 million gas limit floor represents a commitment to L1 capacity that runs counter to the narrative that Ethereum is ceding execution entirely to Layer 2s.
The technical specifications are largely settled. The open variable is time. ePBS implementation is behind schedule, public testnets have no confirmed dates, and the Foundation is undergoing a leadership transition. Whether Glamsterdam ships in Q3 2026 or slips further will test the Ethereum development community's ability to deliver complex protocol changes under a compressed timeline.
The upgrade does not exist in isolation. Hegotá, targeted for late 2026, carries Verkle Trees and FOCIL — items deferred from Glamsterdam to control scope. Vitalik Buterin's March 2026 proposal to eventually replace the EVM with RISC-V looms as a longer-term architectural question. Glamsterdam is the first concrete step in a multi-year execution layer rebuild.