Ethereum's Glamsterdam hard fork — the network's largest protocol overhaul since the September 2022 Merge — entered its final devnet phase on June 16, 2026. The upgrade bundles ten Ethereum Improvement Proposals under meta-EIP 7773, headlined by enshrined Proposer-Builder Separation (EIP-7732) an...
Ethereum's Glamsterdam hard fork — the network's largest protocol overhaul since the September 2022 Merge — entered its final devnet phase on June 16, 2026. The upgrade bundles ten Ethereum Improvement Proposals under meta-EIP 7773, headlined by enshrined Proposer-Builder Separation (EIP-7732) and Block-Level Access Lists (EIP-7928). Together, these changes clear the path for a gas-limit increase from approximately 60 million to 200 million per block, a 233% expansion that targets roughly 10,000 transactions per second under realistic workloads.
No firm mainnet date is set. Current estimates from core developers and Everstake's technical review center on end of August 2026, following public testnet hardening on Holesky and Hoodi. The upgrade's timeline has already slipped from an original H1 2026 target. Past forks have required two to four months of public-testnet validation before mainnet activation, placing the realistic window between September and December 2026.
The economic implications are material. Glamsterdam carries a broad gas repricing that reduces compute-opcode costs while raising state-access charges, a rebalancing that Jayanthi described as: "This will majorly change the cost of actions on Ethereum. High-level compute gets cheaper and state gets more expensive." For Layer-2 rollups, the wider block propagation window expands the data-availability budget, continuing the trajectory set by Fusaka's PeerDAS deployment in December 2025.
The upgrade is tracked under EIP-7773, which reached Draft status on June 17, 2026. It lists ten proposals as Scheduled for Inclusion:
| EIP | Function | |-----|----------| | EIP-7732 | Enshrined Proposer-Builder Separation (ePBS) | | EIP-7928 | Block-Level Access Lists (BALs) | | EIP-7976 | Increases calldata floor cost | | EIP-7981 | Raises EIP-2930 access list costs | | EIP-8037 | Increases state creation gas cost | | EIP-7954 | Raises contract size limit | | EIP-7778 | Removes gas refunds from block accounting | | EIP-8024 | Introduces SWAPN, DUPN, EXCHANGE opcodes | | EIP-7843 | Adds SLOTNUM opcode | | EIP-7708 | Native ETH transfers emit logs |
Two proposals were explicitly excluded. EIP-7782, which would have reduced slot time from 12 seconds to 6 seconds, was deferred due to maturity concerns. FOCIL (Forced Inclusion Lists) was moved to the subsequent Hegotá upgrade to reduce complexity.
Jayanthi characterized the scope: "Probably the largest fork we've had since the Merge."
EIP-7732 — Enshrined Proposer-Builder Separation. Currently, 80-90% of Ethereum blocks are produced using external relays such as MEV-Boost — third-party infrastructure that introduces centralization and trust dependencies. EIP-7732 moves this separation directly into the protocol, replacing off-chain relay infrastructure with in-protocol bids, commitments, and payments.
The technical mechanism extends the data propagation and validation window from approximately 2 seconds to roughly 9 seconds. This is achieved through a new Payload Timeliness Committee (PTC), adding a validator duty that verifies builder payloads are delivered on time. The change touches nearly every layer of the block production path.
The practical effect: validators can outsource block assembly through a protocol-native mechanism rather than trusting external software. The broader network gains stronger censorship resistance by reducing the leverage that a small number of relay operators currently hold over block inclusion.
EIP-7928 — Block-Level Access Lists. BALs function as a dependency map for transaction execution. Each block declares in advance which accounts and storage slots its transactions will access. With this information, clients can identify non-overlapping transactions and execute them in parallel — performing simultaneous disk reads and computations for transactions that touch different parts of the state.
Without BALs, the Ethereum Virtual Machine processes transactions sequentially. With them, the network removes a structural bottleneck that has constrained L1 throughput since genesis. The combination of ePBS (wider propagation window) and BALs (parallel execution) creates the conditions under which validators could safely vote the gas limit up toward 200 million — roughly triple the current ~60 million ceiling.
The 200 million figure is a design target, not a protocol-enforced value. Validators set the limit via standard gas-vote signaling and would step the limit up incrementally only as nodes demonstrate they can handle larger blocks without degraded propagation.
Glamsterdam restructures Ethereum's internal fee schedule. Three EIPs specifically target gas costs:
The net effect is a directional shift: pure computation becomes cheaper; state writes and storage access become more expensive. According to preliminary estimates cited by multiple sources, complex smart contract interactions could see gas costs fall by up to 78% on the compute side, while state-heavy operations absorb higher charges.
This repricing is not cosmetic. It changes the economic calculus for protocol designers, DeFi contract deployers, and rollup sequencers. Applications that perform heavy computation but minimal state changes benefit. Applications that create large amounts of persistent state — such as NFT minting contracts or storage-intensive protocols — face higher per-operation costs.
Glamsterdam's validator impact is concentrated in three areas:
1. New PTC duty. Validators will perform a new attestation duty as part of the Payload Timeliness Committee. This requires running updated consensus and execution layer clients before the fork activates.
2. MEV-Boost dependency reduction. The move of proposer-builder separation on-chain eliminates the current reliance on external relay infrastructure for the ~88% of blocks currently built through MEV-Boost. Relay operators — Flashbots, BloXroute, Ultra Sound, and others — face a structural reduction in their role, though the transition timeline depends on how quickly the ecosystem migrates.
3. Hardware requirements. Node operators need to audit storage IOPS capacity. The combination of larger blocks (if gas limit increases) and parallel execution creates higher disk I/O demands. Operators running on minimal hardware may face performance constraints.
Ethereum's validator set currently stands at approximately 1,239,795 active validators, with 39.6 million ETH staked — 32.61% of the circulating supply. Over the first 165 days of 2026, the network added 96,462 new validators and 4.05 million additional staked ETH, according to data from beaconcha.in. Lido Finance holds 61.66% of the $25.6 billion liquid staking market, with 8.89 million ETH under management.
Glamsterdam continues the trajectory established by Pectra (May 2025) and Fusaka (December 2025) in reducing rollup costs. The wider propagation window enabled by ePBS allows Ethereum to carry more blobs per block, expanding the data-availability budget that rollups draw from.
A typical L2 transaction that cost approximately $0.50 in late 2025 dropped to between $0.20 and $0.30 following Fusaka's PeerDAS deployment. Glamsterdam targets further reductions by moving the blob target toward 72+ per block — a 24x increase from the original EIP-4844 launch parameters.
The calldata repricing (EIP-7976) is specifically designed to push data-posting workloads off the L1 execution layer and into blob space. Rollup sequencers that still post data via calldata — rather than blobs — will face higher costs, accelerating the migration to blob-native data availability.
For L2 users, the practical effect would be reduced periods of congestion-driven fee spikes when multiple rollups compete simultaneously for blob space.
| Milestone | Status / Estimate | |-----------|-------------------| | Soldøgn interop devnet | Concluded May 2, 2026 | | Glamsterdam devnet-5 / devnet-6 | Active as of June 16, 2026 | | EIP-7773 (Meta EIP) | Draft, June 17, 2026 | | Public testnet (Holesky, Hoodi) | Pending — follows devnet completion | | Mainnet activation | Best estimate: end of August 2026 | | Hegotá (subsequent fork) | Tentatively Q4 2026 – Q1 2027 |
Risk factors:
As of mid-June 2026, Ethereum's operating metrics provide context for the upgrade's significance:
| Metric | Value | |--------|-------| | ETH Price | $1,802.77 (June 16, 2026) | | Market Capitalization | $217.4 billion | | 24h Trading Volume | $17.1 billion | | Total Staked ETH | 39.6 million (32.61% of supply) | | Active Validators | ~1,239,795 | | New Validators (Jan–Jun 2026) | +96,462 | | Liquid Staking TVL | $25.6 billion (14.41M ETH) | | Lido Market Share (Liquid Staking) | 61.66% | | Weekly Validator Rewards Issued | 94,525 ETH | | Weekly ETH Burned | 324 ETH | | Annualized Net Inflation | 0.83% | | Current Gas Limit | ~60 million | | Post-Glamsterdam Target Gas Limit | 200 million |
The burn-to-issuance ratio is notably low. The network issued 94,525 ETH in validator rewards over the most recent 7-day period while burning only 324 ETH. This annualized inflation rate of 0.83% contrasts with the "ultrasound money" deflationary thesis that prevailed during high-activity periods in 2023-2024. Glamsterdam's throughput expansion could increase transaction volume and, by extension, base fee burns — but the relationship between capacity expansion and fee revenue is not mechanically guaranteed.
Glamsterdam is in final devnet testing as of June 16, 2026, with ten EIPs locked under meta-EIP 7773. Mainnet activation is estimated for end of August 2026 at earliest, with September-December 2026 as the realistic window.
Two structural changes dominate the upgrade: ePBS (EIP-7732) moves ~88% of block production currently handled through off-chain relays onto the protocol. BALs (EIP-7928) enable parallel transaction execution, unlocking the path to a 200 million gas limit.
Gas repricing shifts economics from cheap-state/expensive-compute toward cheap-compute/expensive-state. Applications relying heavily on state creation face higher costs. Compute-intensive applications benefit.
Validator operations change materially. A new Payload Timeliness Committee duty is added. All consensus and execution clients must be updated before activation. Hardware requirements — particularly disk IOPS — may increase with larger blocks.
L2 rollups gain expanded blob budgets, continuing cost reductions begun by Fusaka's PeerDAS deployment. The calldata repricing further incentivizes migration to blob-native data availability.
No mainnet date is confirmed. The fork's scope — described by core developers as the largest since the Merge — and its prior schedule slip introduce non-trivial execution risk.
Glamsterdam represents the most technically ambitious Ethereum upgrade since the network transitioned to proof-of-stake in September 2022. The combination of enshrined proposer-builder separation, parallel execution via block-level access lists, and comprehensive gas repricing addresses three distinct bottlenecks: MEV infrastructure centralization, sequential execution throughput, and misaligned fee incentives.
The practical question is execution. Ten EIPs affecting both consensus and execution layers must pass through testnet hardening without introducing regressions. The timeline has already slipped once. Validators, node operators, and contract developers face meaningful preparation requirements before activation.
If delivered on the current trajectory, Glamsterdam positions Ethereum's L1 to process roughly 3x its current transaction volume while reducing its dependency on off-chain relay infrastructure. Whether that translates into increased economic activity — and by extension, increased fee revenue and ETH burns — depends on whether demand materializes to fill the expanded capacity. The upgrade provides the supply side. The demand side remains an open variable.