Ethereum's Glamsterdam upgrade — ten EIPs tracked under meta-specification EIP-7773 — entered public testing on August 20, 2026 with the launch of the Platåberget testnet. The Sepolia fork is provisionally scheduled for September 28 at 14:44:48 UTC (epoch 351,232, slot 11,239,424), pending confir...
"At their best, interop weeks can compress a month of asynchronous progress into each day." — Tim Beiko, Ethereum Foundation Researcher
Ethereum's Glamsterdam upgrade — ten EIPs tracked under meta-specification EIP-7773 — entered public testing on August 20, 2026 with the launch of the Platåberget testnet. The Sepolia fork is provisionally scheduled for September 28 at 14:44:48 UTC (epoch 351,232, slot 11,239,424), pending confirmation at ACDC Call #186 on September 3. No mainnet date has been set. Circulating projections of November 4 lack official approval.
Glamsterdam ships two structural changes to Ethereum's architecture. EIP-7732 enshrines Proposer-Builder Separation (ePBS) directly into the consensus protocol, eliminating reliance on off-chain relay infrastructure through which over 88% of blocks are currently routed. EIP-7928 introduces Block-Level Access Lists (BALs), enabling parallel transaction execution and clearing a path for the gas limit to rise from 60 million toward a 200-million floor — roughly tripling Layer 1 capacity.
The upgrade is the largest protocol restructuring since the Merge in September 2022. It arrives at a moment when Ethereum's economic model faces sustained pressure: ETH remains down over 50% from its August 2025 peak of $4,946, and Layer 1 fee revenue has compressed as activity migrates to rollups. Whether Glamsterdam reverses that trajectory depends on whether expanded L1 capacity generates proportional demand — a question the data cannot yet answer.
The Ethereum Foundation announced Platåberget on August 17, 2026, with the Glamsterdam fork activating on the testnet three days later. The name references the mountain near Longyearbyen, Svalbard, where just over 100 core contributors gathered for the interop sprint that finalized gas limit parameters, stabilized external block builder implementations, and locked gas repricing numbers for EIP-8037.
Platåberget is a purpose-built, short-lived testnet designed to run for several months. It precedes deployment to Ethereum's longer-lived testnets — Sepolia and Hoodi — and serves as the first environment where external validators and builders can interact with the full Glamsterdam feature set. Public deposits are open; anyone can run validators and builders using container images provided by ethPandaOps for six consensus clients (Lighthouse, Lodestar, Nimbus, Prysm, Teku, Grandine) and seven execution clients (Besu, Geth, Erigon, Nethermind, Reth, Nimbus-EL, Ethrex).
The Foundation's blog post flagged a critical breaking change: any tool that relies on a hardcapped maximum gas limit — wallets, indexers, gas estimators — will break and requires updating. A new state gas dimension (EIP-8037) creates separate costs for state-touching operations, meaning plain ETH transfers to new accounts will incur gas costs beyond the traditional 21,000-gas baseline.
Today, Ethereum's block production is divided informally. Validators (proposers) select the block, and specialized builders assemble the execution payload containing transactions. This separation exists almost entirely through MEV-Boost, an off-chain relay system developed by Flashbots. According to Everstake, over 88% of Ethereum blocks are currently built through this off-chain pathway.
EIP-7732 moves the proposer-builder handoff into the consensus protocol itself. Under ePBS, the proposer commits to a block header, and a separate builder constructs and reveals the execution payload. A new validator duty — the Payload Timeliness Committee (PTC) — attests to whether the committed payload was revealed correctly and on time.
The technical consequence is significant: the payload propagation window expands from approximately 2 seconds to roughly 9 seconds. This wider window unblocks Ethereum's ability to handle higher transaction throughput without compromising block propagation safety. It also eliminates the trust assumption embedded in relay infrastructure — proposers no longer need to trust a third-party relay to faithfully deliver the builder's payload.
The MEV implications are material. Proponents estimate ePBS could reduce MEV centralization by up to 70%, making block-building rewards more accessible to smaller, home-staked validators. However, capital-intensive builders with sophisticated infrastructure may still dominate block construction through superior order-flow capture. ePBS changes the plumbing, not the economics of information asymmetry.
The EIP's status within EIP-7773 remains "Seeking Further Input" (SFI) — not yet formally locked for inclusion. Ethereum Foundation contributors have stated that getting ePBS right outranks any fixed date.
EIP-7928 introduces Block-Level Access Lists (BALs) — metadata that records which state locations each transaction accesses, transmitted separately from block bodies via a new eth/71 wire protocol message. This gives validators instant visibility into which transactions do not overlap, enabling three capabilities:
The practical outcome: BALs clear the path for Ethereum's gas limit to rise from its current 60 million to a 200-million floor. At 200 million gas per block, Ethereum proponents project approximately 10,000 TPS-equivalent throughput under realistic workloads — roughly a 3x capacity increase from today's L1 limits.
This is not automatic. The gas limit is set by validators through a soft governance mechanism (signaling their preferred limit), not by the hard fork itself. Glamsterdam makes the higher limit technically safe; whether validators signal to raise it is a separate coordination exercise.
Glamsterdam includes a package of gas repricing EIPs that realign operation costs with actual computational expense on modern hardware. Many current gas prices were calibrated years ago and no longer reflect real resource consumption.
Key fee-related changes:
The projected fee impact: approximately 78.6% reduction in L1 fees for common transaction types, according to modeling cited by multiple sources. However, this projection assumes constant demand. If L1 usage scales with capacity — which is the stated goal — the base fee may remain stable or increase. Gas fees are ultimately demand-driven.
Glamsterdam ships ten EIPs under the EIP-7773 meta specification. Beyond the three headliners:
Notable exclusions: EIP-7782 (6-second slots) was declined; it would have delayed ZK-proving and forced timing recalibration across the stack. FOCIL (Fork-Choice Inclusion Lists) was moved to the subsequent Hegotá upgrade (targeted H1 2027) to reduce implementation complexity.
The current schedule, as discussed at the August 20 ACDC call:
| Milestone | Date | Status | |-----------|------|--------| | Platåberget testnet fork | August 20, 2026 | Completed | | ACDC Call #186 (confirm Sepolia date) | September 3, 2026 | Pending | | Sepolia testnet fork | September 28, 2026 | Provisional | | Hoodi testnet fork | October 26, 2026 | Provisional, contingent on stable DevNets | | Mainnet activation | TBD (Q4 2026 target) | No date set |
Dima Gusakov of Lido requested "at least one day and preferably two on a stable DevNet" for oracle and tooling testing, and preferred "both dates a bit later — like a week or so."
DevNet 8 experienced a finality loss caused by a builder-deposit caching bug affecting Lighthouse, Prysm, and Teku clients. The bug has been resolved, but it illustrates the implementation risk inherent in ePBS — the most complex consensus-layer change since the Merge.
The EIP-7773 meta specification remains in Draft status. Ethereum Foundation contributors have consistently emphasized that correctness takes precedence over schedule. A slip into late Q4 or beyond remains plausible if cross-client feature parity is not achieved before the Sepolia fork window.
For institutional validators and staking providers, Glamsterdam changes operational requirements in several areas:
Client updates are mandatory. Both consensus-layer clients (Lighthouse, Prysm, Teku, Nimbus, Lodestar) and execution-layer clients (Geth, Nethermind, Besu, Erigon) must be updated before mainnet activation. Validators running outdated clients will fork off the network.
Exit liquidity improves materially. EIP-8061's 4.8x faster exit processing reduces one of staking's most operationally consequential risks — the cost associated with longer time-to-liquidity. For large institutional positions, this compresses the window during which capital is locked during an exit.
MEV-Boost dependency shifts. ePBS does not immediately deprecate MEV-Boost, but it creates a protocol-native alternative. Institutional operators must evaluate whether to adopt the new in-protocol builder market or maintain existing relay relationships. Opting into the new market is voluntary.
Hardware requirements remain stable. Parallel execution under BALs requires node operators to audit storage IOPS, but current home-validator hardware is deemed "largely sufficient" according to Everstake. The upgrade does not impose new minimum hardware thresholds.
Figment and Everstake have both confirmed plans to support the upgrade across all validator infrastructure.
Glamsterdam represents Ethereum's most ambitious protocol restructuring since the Merge. The upgrade attempts to simultaneously solve three distinct problems: MEV centralization through ePBS, L1 throughput constraints through BALs and gas limit expansion, and staker operational friction through faster exit processing.
The scope is both the upgrade's strength and its primary risk. ePBS alone — moving block building into the consensus layer — constitutes a fundamental change to how Ethereum's block production works, touching every consensus client and every builder in the ecosystem. Adding parallel execution infrastructure and gas repricing on top creates a testing surface that exceeds any prior Ethereum hard fork.
The Platåberget testnet's August 20 launch marks the beginning, not the end, of the validation process. The September 3 ACDC call will determine whether the Sepolia timeline holds. From an economic-value perspective, Glamsterdam's impact on Ethereum's fee-revenue trajectory — and by extension, ETH's value proposition as a productive asset — depends on whether 3x L1 capacity generates proportional transaction demand or simply compresses per-transaction revenue further. The protocol is betting on the former. The data will follow.