Ethereum's Glamsterdam hard fork — a portmanteau of the Gloas consensus-layer and Amsterdam execution-layer upgrades — targets mainnet activation at the end of August 2026, pending final testnet validation. It is the network's most substantial protocol change since the September 2022 Merge. The u...
"This is probably the largest fork we've had since the Merge." — Parithosh Jayanthi, Ethereum Foundation
Ethereum's Glamsterdam hard fork — a portmanteau of the Gloas consensus-layer and Amsterdam execution-layer upgrades — targets mainnet activation at the end of August 2026, pending final testnet validation. It is the network's most substantial protocol change since the September 2022 Merge. The upgrade bundles ten Ethereum Improvement Proposals (EIPs), headlined by EIP-7732 (Enshrined Proposer-Builder Separation) and EIP-7928 (Block-Level Access Lists), which together restructure how blocks are built and executed on the base layer.
The economic implications are significant. The upgrade's gas repricing package is projected to cut certain L1 transaction fees by approximately 78%, while ePBS is designed to reduce MEV extraction by up to 70%. The block gas limit is targeted to rise from 60 million to 200 million, pushing theoretical throughput toward 10,000 transactions per second — roughly a 10x increase from current capacity. These changes represent a strategic pivot: after years of deferring scaling to Layer 2 rollups, Ethereum is rebuilding the base layer as a high-performance settlement and execution engine.
Glamsterdam is governed by Meta EIP-7773, which as of June 2026 lists ten proposals as Scheduled for Inclusion (SFI):
| EIP | Name | Layer | Function | |------|------|-------|----------| | 7732 | Enshrined PBS (ePBS) | Consensus | Moves proposer-builder separation into the protocol | | 7928 | Block-Level Access Lists | Execution | Enables parallel transaction execution | | 7976 | Increase Calldata Floor Cost | Execution | Gas repricing — raises calldata costs | | 7981 | Increase Access List Cost | Execution | Gas repricing — raises access list costs | | 8037 | State Creation Gas Cost Increase | Execution | Gas repricing — raises state creation costs | | 7954 | Increase Maximum Contract Size | Execution | Raises contract deployment size ceiling | | 7778 | Block Gas Accounting Without Refunds | Execution | Simplifies gas accounting model | | 8024 | Backward-Compatible SWAPN, DUPN, EXCHANGE | Execution | EVM opcode improvements | | 7843 | SLOTNUM Opcode | Execution | Adds slot number to EVM context | | 7708 | ETH Transfers Emit a Log | Execution | Native ETH transfers produce event logs |
Two additional proposals — EIP-7688 (forward-compatible consensus data structures) and EIP-8061 (increased exit and consolidation churn) — hold Considered for Inclusion (CFI) status, according to Figment staff analyst Benjamin Thalman. EIP-7782 (6-second slots) was deferred from Glamsterdam, and FOCIL (Fork-Choice Inclusion Lists) was pushed to the subsequent Hegotá fork.
The scope is not yet final. EIP-7773 remains in draft, and the proposal set could narrow or expand before mainnet activation.
Today, approximately 88% of Ethereum blocks are built through MEV-Boost, an off-protocol relay system maintained by Flashbots and a handful of other operators, according to Everstake. Only seven of eleven active relays hold more than 1% market share. The system works but depends on trusted third-party infrastructure that the protocol cannot verify or enforce.
EIP-7732 moves the proposer-builder handoff directly into the consensus layer. Under ePBS, block proposers (validators selected to propose a slot) accept a payment from a builder without needing to see the full execution payload upfront. A new Payload Timeliness Committee (PTC) validates that the builder delivered the payload on time. The data propagation window extends from approximately 2 seconds to roughly 9 seconds, accommodating the two-phase commit.
The practical effects:
For institutional stakers, the change alters validator economics. Relay tips — a revenue source dependent on MEV-Boost configuration — are replaced by protocol-native builder payments. The fee structure becomes more transparent but potentially less variable.
Block-Level Access Lists (BALs) function as a pre-execution index. Each block declares which accounts and storage slots will be accessed, along with their post-execution values. This declaration enables four forms of parallelism:
The BAL size is constrained by the block gas limit, not a fixed item count. Combined with the other EIPs in the gas repricing package, BALs provide the safety margin to increase the block gas limit from the current 60 million to 200 million — more than tripling current capacity.
For context: Ethereum's gas limit sat at approximately 30 million from 2020 through January 2025. It was raised incrementally during 2025 — 36 million by February, 45 million by July, 60 million by November. The Glamsterdam target of 200 million would represent a 233% increase from current levels and a 567% increase from the 2020-2024 baseline.
Three EIPs — 7976, 7981, and 8037 — form a gas repricing package that restructures the cost model. The direction, as described by Ethereum Foundation developer Parithosh Jayanthi: "high-level compute gets cheaper and state gets more expensive."
The net effect is projected to cut L1 fees by approximately 78% for certain transaction types, though the exact impact varies by transaction composition and network load. If fees stabilize below $0.50 per transaction, as some projections suggest, Ethereum L1 becomes cost-competitive with L2 rollups for many transaction categories.
This creates a paradox for EIP-1559 economics. Lower per-transaction fees reduce the ETH burn rate. Daily L1 fee revenue has already declined from near $30 million in 2021-2022 to single-digit millions for much of 2026. A threefold increase in capacity without a proportional increase in demand could suppress the burn further, weakening the deflationary mechanism that ETH's monetary thesis relies upon.
The counterargument: cheaper fees attract activity that was previously priced out of L1, potentially increasing total volume enough to offset the per-transaction reduction. This remains unproven.
The current state of development, as of August 2026:
Three risks could delay activation into Q4 2026 or later:
The end-of-August 2026 target is described as "aspirational" by Ethereum Foundation sources. Given that recent hard forks have required two to four months of public-testnet running before mainnet activation, a September-to-December 2026 window represents the firmer base case.
Glamsterdam's core strategic objective is to shift Ethereum's scaling narrative from L2-only throughput back toward a high-performance L1. This represents a reversal of the 2023-2025 orthodoxy that treated the base layer primarily as a settlement and data availability layer.
The revenue dynamics are stark. Base, Coinbase's L2, reportedly generated over $75 million in 2025 while paying Ethereum approximately $1.52 million in settlement fees — a 98% margin for the L2. As activity migrated to rollups, mainnet fee revenue and the EIP-1559 burn fell to levels that undermine Ethereum's security-budget model.
Vitalik Buterin publicly questioned the rollup-only approach in early 2026, noting that only a handful of major L2s had reached meaningful decentralization. He termed the "Lean Ethereum" roadmap — which encompasses Glamsterdam, Hegotá, and subsequent upgrades — as the network's "third major iteration," comparable in scale to the Merge.
The open question: if Glamsterdam cuts gas by 78% and reduces MEV by up to 70%, does activity flow back to L1, where security guarantees are structurally higher than any rollup? Or does L2 inertia — established user bases, application-specific optimization, and lower latency — keep the majority of activity on rollups regardless of L1 fee parity?
At current capacity, Ethereum L1 handles 15-30 transactions per second. Solana advertises 50,000+ TPS theoretical capacity, though sustained real-world throughput under load typically falls in the 3,000-5,000 TPS range.
At 10,000 TPS, post-Glamsterdam Ethereum would narrow this gap materially. While Solana would retain a throughput advantage at peak theoretical capacity, the effective real-world difference shrinks. Ethereum maintains advantages in other dimensions: over $70 billion in validator collateral (approximately 30% of supply staked), more than 1 million validators, and the deepest DeFi ecosystem by TVL.
The competitive frame shifts from "Ethereum is slow and expensive, Solana is fast and cheap" to a comparison of tradeoffs in decentralization, security budget, and ecosystem depth at roughly comparable L1 price points. Solana's own upgrade path — the Alpenglow consensus overhaul targeting 150ms finality — ensures the throughput race continues.
Glamsterdam restructures two of Ethereum's most fundamental mechanisms — block production and transaction execution — in a single hard fork. If delivered as specified, it would represent the most consequential protocol change since proof-of-stake activation in 2022. The 200 million gas limit target alone would increase raw capacity by 3.3x.
The economic implications are double-edged. Cheaper fees and reduced MEV benefit users and may attract activity back to L1. But the same fee reductions threaten the burn mechanism that underpins ETH's scarcity argument. Whether Glamsterdam creates a net-positive value loop — lower fees driving higher volume driving sustainable burns — or a net-negative one — lower fees driving lower burns driving weaker monetary premium — depends on demand elasticity that no model has reliably predicted.
The fork's readiness clock is ticking. Nine client teams, ten EIPs, and public testnets stand between the current devnet and mainnet activation. History suggests the process takes longer than developers hope.