Ethereum's Glamsterdam hard fork — tracked under meta-EIP-7773 — entered its final devnet phase in June 2026 and targets mainnet activation in late Q3 2026, likely end of August. The upgrade bundles ten Ethereum Improvement Proposals that collectively represent the largest structural change to th...
"Probably the largest fork we've had since the Merge. This will majorly change the cost of actions on Ethereum. High-level compute gets cheaper and state gets more expensive." — Parithosh Jayanthi, Core Developer, Ethereum Foundation
Ethereum's Glamsterdam hard fork — tracked under meta-EIP-7773 — entered its final devnet phase in June 2026 and targets mainnet activation in late Q3 2026, likely end of August. The upgrade bundles ten Ethereum Improvement Proposals that collectively represent the largest structural change to the protocol since the September 2022 Merge.
Two proposals carry the bulk of the architectural weight. EIP-7732 enshrines Proposer-Builder Separation (ePBS) directly into the consensus layer, eliminating the third-party relay infrastructure that currently intermediates over 90% of Ethereum block construction. EIP-7928 introduces Block-Level Access Lists (BALs), enabling nodes to identify non-conflicting transactions and process them in parallel. Together, these changes clear the path for a validator-signaled gas-limit increase from the current 60 million to a target of 200 million per block — a 3.3x capacity expansion — and a throughput target of approximately 10,000 transactions per second, roughly ten times the current rate.
A supporting gas-repricing package (EIP-7904) realigns opcode costs with modern hardware benchmarks, projecting a ~78.6% reduction in gas costs for both simple ETH transfers and complex smart-contract interactions. The Ethereum Foundation's DevOps team has tested three EIPs on Devnet-4 and is transitioning to Devnet-5. No mainnet date is locked.
EIP-7773, the Glamsterdam meta-EIP, bundles ten proposals as Scheduled for Inclusion as of June 17, 2026:
| EIP | Description | |-----|-------------| | EIP-7708 | ETH transfers and burns emit a log event | | EIP-7732 | Enshrined Proposer-Builder Separation (ePBS) | | EIP-7778 | Block gas accounting without refunds | | EIP-7843 | SLOTNUM opcode | | EIP-7928 | Block-Level Access Lists (BALs) | | EIP-7954 | Maximum contract size raised from ~24 KiB to 32 KiB | | EIP-7976 | eth/70 partial block receipt lists | | EIP-8024 | Backward-compatible SWAPN, DUPN, EXCHANGE opcodes | | EIP-8037 | State-creation gas-cost increase | | EIP-8159 | eth/71 Block Access List Exchange |
The two headliners — EIP-7732 and EIP-7928 — address block production and execution, respectively. The remaining eight handle gas accounting, EVM opcode extensions, state-cost adjustments, and networking protocol upgrades. The package is designed as a cohesive unit: ePBS extends the data propagation window, which BALs then exploit for parallel processing, which in turn makes the 200M gas limit feasible.
Since the Merge, Ethereum's block-production pipeline has relied on MEV-Boost, a third-party middleware system developed by Flashbots. Under the current architecture, validators (proposers) outsource block construction to specialized builders via relays. According to relay data from relayscan.io as of July 14, 2026, the relay landscape breaks down as follows:
On the builder side, concentration is more severe. Two builders — Titan (52.60% of blocks) and Quasar (21.63%) — control over 74% of block production as of January 2026 data. Five builders collectively account for 96.7% of all MEV blocks.
This concentration creates a systemic dependency on off-protocol infrastructure. If relays go offline or collude, block production degrades. The Aestus relay alone serves over 650,000 validators, according to its operational disclosures.
EIP-7732 moves the proposer-builder handoff into the consensus protocol itself. Under ePBS:
This eliminates the need for trusted relays entirely. Research from the Flashbots Collective estimates that ePBS could reduce MEV-related losses for ordinary users by up to 70%, according to thirdweb's technical analysis.
A critical secondary effect: ePBS extends the execution-payload propagation window from approximately 2 seconds to roughly 9 seconds. This 4.5x expansion is what enables larger blocks and higher blob counts in subsequent forks.
Current Ethereum execution is sequential: the EVM processes transactions one after another, each waiting for the prior state update. EIP-7928 introduces Block-Level Access Lists that declare which state keys each transaction will read or write. Nodes use this metadata to identify non-conflicting transactions and execute them simultaneously.
The mechanism works as follows:
This is not speculative parallelism (which Solana and Monad use). BALs provide deterministic parallelism — the access lists guarantee non-conflict before execution begins.
The practical throughput target is approximately 10,000 TPS on Layer 1, roughly a 10x increase from the current ~1,000 TPS effective rate. This figure assumes the gas limit reaches 200 million and nodes demonstrate stable propagation at that block size.
Many of Ethereum's gas costs were calibrated years ago and no longer reflect the computational expense on modern hardware. EIP-7904 realigns opcode pricing with current execution benchmarks.
As Parithosh Jayanthi of the Ethereum Foundation stated: "High-level compute gets cheaper and state gets more expensive."
The 200M figure is a design target, not a protocol-enforced parameter. Validators set the gas limit via standard gas-vote signaling. They currently coordinate around 60 million. The path to 200 million requires:
This graduated approach mirrors how the gas limit rose from 30 million to 60 million over the course of 2024–2025. The timeline from 60M to 200M is unspecified and depends on empirical network performance.
Glamsterdam does not directly increase blob count — that was primarily a Fusaka/PeerDAS outcome. However, it enables future blob expansion through two mechanisms:
Longer propagation windows. The ePBS-driven expansion from 2 to 9 seconds of payload propagation time gives the network more room to handle larger execution payloads alongside blob data. Future forks can safely increase blob targets without competing for the same narrow propagation window.
L1 fee reduction. With L1 transactions becoming 71–78% cheaper, the cost calculus for L1 vs. L2 shifts. Applications that currently use rollups primarily for fee savings may reconsider L1 deployment. This does not threaten L2 viability — rollups still offer higher throughput and application-specific customization — but it narrows the fee differential that has been L2s' primary marketing argument.
For rollup operators, the more significant development is the larger block capacity. More L1 gas headroom means more calldata and blob space can be priced into blocks, reducing the base cost of posting L2 state roots and proofs.
As of mid-2026, Ethereum's staking landscape includes approximately 1.24 million active validators with 38.9 million ETH staked (~32% of circulating supply), according to beaconcha.in. Staking yields sit at 3.0–3.5% across major venues.
Glamsterdam affects validators in several ways:
Reduced operational complexity. ePBS eliminates the need to run MEV-Boost sidecar software. Validators currently must maintain relay connections, monitor builder bids, and handle fallback local block production. Post-Glamsterdam, the protocol handles this natively.
More level reward distribution. Empirical data from ePBS research indicates more uniform block-reward distribution across proposers. Under the current system, validators connected to high-performance relays capture disproportionate MEV. Protocol-level PBS reduces this variance.
Higher hardware requirements. The 200M gas limit — if and when validators vote it in — requires nodes to process and propagate 3.3x larger blocks. Node operators will need to evaluate storage I/O, bandwidth, and CPU capacity. This is a medium-term concern, not an immediate fork-day requirement.
Consolidation dynamics continue. The Pectra upgrade (May 2025) raised the maximum effective balance from 32 ETH to 2,048 ETH. Validator consolidation remains ongoing but slower than expected — the economic boost from native compounding is modest for large operators, while operational and risk costs of merging keys are non-trivial. Lido's Staking Router v3 hit mainnet in July 2026, enabling balance-based accounting for 0x02 validators.
| Milestone | Status / Target | |-----------|----------------| | EIP-7773 Meta EIP | Draft (June 17, 2026) | | Devnet-4 | Three EIPs tested; transitioning to Devnet-5 | | Final devnet phase | Entered June 2026 | | Holesky testnet | Pending — estimated 14–28 days runtime | | Sepolia testnet | Pending — estimated 7–14 days after Holesky | | Mainnet activation | Target: late August 2026 (aspirational) | | Firm base case | September–December 2026 |
The minimum critical path from audit completion to mainnet requires approximately 65–93 days: 30 days for audit, 14–28 days on Holesky, 7–14 days on Sepolia, and 14–21 days of mainnet lead time.
Parithosh Jayanthi described the current phase as: "Mostly testing, finalizing specs, outreach to the community about repricings and then shipping it."
The subsequent fork, Hegota, is expected later in 2026 or early 2027 and will build on the infrastructure Glamsterdam establishes.
Glamsterdam represents Ethereum's pivot from scaling exclusively through rollups to scaling the base layer itself. The combination of ePBS and BALs addresses two longstanding criticisms: centralized block construction and sequential execution. If the upgrade ships as designed, Ethereum's L1 will offer meaningfully cheaper transactions, higher throughput, and reduced MEV extraction — narrowing the gap with competing L1s that have marketed these features for years.
The risks are proportional to the ambition. Ten EIPs activating simultaneously create a large attack surface for bugs. The 200M gas limit is an engineering aspiration, not a guaranteed outcome. And ePBS introduces new game-theoretic dynamics — concerns about multiparty collusion and chain stagnation are documented in academic literature (see arxiv.org/abs/2506.18189).
The market will judge Glamsterdam on execution, not promises. The testnet deployments on Holesky and Sepolia in the coming weeks will provide the first empirical data on whether these design-stage targets survive contact with production traffic.