Ethereum's Glamsterdam hard fork — a portmanteau of the consensus-layer "Gloas" and execution-layer "Amsterdam" upgrades — entered unified devnet testing in late April 2026. The upgrade targets a 3.3x gas limit increase (60 million to 200 million per block), an estimated 78.6% reduction in execut...
"We don't friggin need more copypasta EVM chains." — Vitalik Buterin, Ethereum Co-Founder, February 2026
Ethereum's Glamsterdam hard fork — a portmanteau of the consensus-layer "Gloas" and execution-layer "Amsterdam" upgrades — entered unified devnet testing in late April 2026. The upgrade targets a 3.3x gas limit increase (60 million to 200 million per block), an estimated 78.6% reduction in execution gas costs, and a throughput ceiling of approximately 10,000 transactions per second on the base layer. If activated on schedule in Q3 2026, it would represent the third hard fork in twelve months, following Pectra (May 2025) and Fusaka (December 2025).
Two structural changes anchor the upgrade: EIP-7732 (Enshrined Proposer-Builder Separation) and EIP-7928 (Block-Level Access Lists). The first replaces Ethereum's off-chain block-building relay market with an in-protocol commit-reveal mechanism. The second enables parallel transaction execution by declaring storage-slot dependencies at the block level. Together, they address a problem that has concentrated block production among two builders controlling over 75% of blocks and routed more than 50% of high-value transactions through private channels to avoid MEV extraction.
The stakes are significant. With over 85% of Ethereum blocks currently produced through MEV-Boost relays — off-chain infrastructure that was always intended as temporary — Glamsterdam's ePBS represents the protocol's first attempt to reclaim block construction as a native function. Whether it succeeds without introducing new centralization vectors remains an open question.
The Soldøgn Interop event, held in Svalbard, Norway, concluded on May 2, 2026. During the event, developers ran all major Glamsterdam components together in a unified test environment for the first time. The interop aimed to synchronize eight different client implementations by Monday; actual synchronization was achieved by Tuesday, with the remainder of the week spent on stress testing.
Prior to the interop, the Ethereum Foundation DevOps team completed EIP testing on Devnet-4 and has since transitioned to Devnet-5. According to the Ethereum Foundation's Checkpoint #9, published April 10, 2026, development is proceeding "slowly but steadily," with ePBS identified as the primary bottleneck due to cross-client coordination complexity.
The timeline:
| Milestone | Status | |-----------|--------| | Separate ePBS and BALs devnets | Completed (Q1 2026) | | Devnet-4 (individual EIP testing) | Completed (April 2026) | | Soldøgn Interop (unified testing) | Completed (May 2, 2026) | | Devnet-5 (combined features) | In progress | | Holesky/Sepolia public testnets | Pending (spring/summer 2026) | | Dual audit rounds | Pending | | Mainnet activation | Target: Q3 2026 |
Tomasz Stańczak, co-director of the Ethereum Foundation, stated at the Bankless Summit that the gas limit would increase to 100 million in H1 2026, with a doubling to 200 million following ePBS activation. The original aspirational target of June 2026 for mainnet activation has effectively slipped. Q3 2026 is now the consensus estimate among core developers, pending successful testnet validation.
Ethereum's current block production relies on MEV-Boost, an off-chain relay system introduced after the Merge in September 2022 as a stopgap measure. The system separates block proposers (validators) from block builders (specialized entities that construct optimally ordered blocks). As of April 2026, over 85% of Ethereum blocks are produced through MEV-Boost relays. The relay market itself shows concentration: Ultrasound Money handles 33.9% of relay payloads, Titan 24.2%, and BloXroute 14.7%.
An estimated 80-90% of Ethereum block production currently depends on off-chain builders operating within trust-based agreements with relays. By mid-2025, more than 50% of high-value Ethereum transactions were routed through private channels to circumvent MEV extraction. Cumulative MEV extracted across Ethereum surpassed $1.8 billion by mid-2025, with $40-60 million captured per month, concentrated in arbitrage and sandwich strategies.
EIP-7732 moves the proposer-builder market on-chain through a commit-reveal mechanism:
The propagation window expands from approximately 2 seconds to 9 seconds, giving builders more time to optimize blocks without creating timing-game incentives.
The design introduces what researchers term the "free option problem." In approximately 0.82% of blocks under normal conditions — rising to 6% during high volatility — builders may gain an asymmetric advantage by observing market movements between commitment and revelation. Additionally, while ePBS decentralizes the proposer layer, it may inadvertently increase builder consolidation by raising the technical sophistication required to compete.
EIP-7928 introduces a data structure called Block-Level Access Lists (BALs) that records every account and storage slot accessed during block execution. Each block header gains a new field: block_access_list_root.
The practical effect: nodes can prefetch the entire working state before execution begins, and non-conflicting transactions can be processed in parallel across multiple threads. Current Ethereum execution is strictly sequential — each transaction must complete before the next begins.
Supporting infrastructure changes include:
BALs also have downstream implications for zero-knowledge proof generation. Explicit state-access records simplify ZK verification by eliminating the need for mid-execution storage lookups, a step toward the Ethereum Foundation's longer-term ZK-EVM roadmap.
The gas limit has already doubled once since the Merge: from 30 million to 60 million across the Pectra and Fusaka upgrades in 2025. Glamsterdam targets a phased increase:
The raw gas-limit increase is complemented by EIP-7904 (General Gas Repricing), which reprices EVM opcodes based on empirical hardware benchmarks rather than historical pricing. The combined effect targets a 78.6% reduction in gas costs for both simple ETH transfers (where EIP-2780 alone could cut intrinsic gas by up to 71%) and complex smart contract interactions.
At current gas prices, users paying $4-8 per swap on Ethereum L1 would see those costs drop to approximately $1-2 under Glamsterdam pricing. Throughput scales from an effective ~1,000 TPS to a theoretical ceiling of 10,000 TPS, though real-world throughput will depend on actual transaction complexity and BAL adoption.
Two additional gas-related EIPs manage the sustainability side:
This creates a deliberate asymmetry: executing existing contract logic gets cheaper; creating new permanent state gets more expensive.
Beyond the two headline proposals, the Glamsterdam fork includes up to 22 EIPs. Notable inclusions:
| EIP | Description | |-----|-------------| | EIP-8045 | Excludes slashed validators from proposing future blocks | | EIP-8080 | Enables exit consolidation queuing, up to 2.5x faster exits during high demand | | EIP-7997 | Deterministic factory predeploy for identical contract addresses across EVM chains | | EIP-7708 | ETH transfers and burns emit event logs for tracking | | EIP-7975 (eth/70) | Partial block receipt lists to prevent sync failures |
Over 30 proposals were explicitly declined, including FOCIL (Forced Inclusion Lists), reduced slot times, multidimensional gas metering, and post-quantum signature verification. FOCIL, which would have required randomly selected validators to enforce transaction inclusion for censorship resistance, was deferred despite Vitalik Buterin publicly advocating for "big FOCIL" and encrypted mempools as long-term solutions to block-building centralization. The decision to exclude it from Glamsterdam reflects a scope-management trade-off.
The economic rationale for ePBS is rooted in observable market structure deterioration. Key data points:
The concentration stems from exclusive private order flow agreements. Applications sell transaction order flow directly to builders, bypassing the public mempool. According to academic research analyzing January 2023 to May 2024, private order flows contributed to 54.6% of total block value.
From an economic value distribution perspective, this architecture extracts rent from end users and redirects it to a narrow set of infrastructure operators — precisely the type of value-chain inefficiency that protocol-level reform can address. ePBS does not eliminate MEV, but it standardizes the handoff rules and removes the trust dependency on relay operators who currently serve as unregulated intermediaries.
Implementation complexity. Neither ePBS nor BALs has been tested at mainnet scale. The consensus-layer changes required by ePBS are substantial — the PTC mechanism adds a new committee structure to the beacon chain, and the commit-reveal flow introduces novel timing assumptions.
Builder consolidation paradox. ePBS may reduce relay concentration while increasing builder concentration. The technical requirements for competitive block building — low-latency data feeds, cross-chain arbitrage infrastructure, exclusive order flow relationships — remain unchanged. Moving the market on-chain does not flatten the skill gap.
Gas limit sustainability. A 200 million gas limit at current storage growth patterns could push state database size past 100 GiB per year without the countervailing effects of EIP-8037 and EIP-8038. The effectiveness of the "cheaper execution, more expensive state creation" asymmetry is untested.
FOCIL absence. Without forced inclusion lists, Glamsterdam's ePBS does not directly address censorship resistance at the builder level. A dominant builder can still exclude transactions, even if the mechanism by which they gain dominance is now protocol-governed.
L2 competitive dynamics. A 78% reduction in L1 gas costs narrows the cost advantage of Layer 2 solutions. Rollups, which derive their value proposition partly from cheaper execution, may face margin compression if L1 becomes sufficiently affordable for routine transactions.
Glamsterdam represents Ethereum's most structurally significant upgrade since the Merge. Where Pectra and Fusaka delivered incremental parameter changes — staking limits, blob capacity — Glamsterdam rewires how blocks are built and executed. The upgrade's two headline EIPs address a real and documented market failure: the concentration of block production in off-chain infrastructure controlled by a small number of entities.
The economic implications extend beyond gas fees. By moving block construction on-chain, Ethereum internalizes a value flow that currently accrues to relay operators and exclusive-order-flow intermediaries. Whether this redistribution benefits end users or simply shifts rent extraction to a different set of sophisticated actors depends on implementation details that remain in flux.
The testing timeline is tight. From Devnet-5 to public testnets to dual audits to mainnet activation in Q3 — each phase compresses further as scope expands. The Ethereum Foundation's stated priority — correctness over speed — suggests additional delays are possible. For now, the protocol is in engineering execution mode, and the data from Soldøgn Interop suggests the core components work. Whether they work at scale is the next question.