Ethereum's Glamsterdam hard fork activated on the Sepolia testnet on October 6, 2026, at 13:53:36 UTC, tripling the block gas limit from 60 million to 200 million. Early test blocks processed between 52 million and 92 million gas — 26% to 46% of the new ceiling — while 99.97% of eligible stake pa...
"Any teenager with disposable test ETH could stall the chain." — Ethereum Core Developer, on builder abuse risk in Glamsterdam testnet (CoinDesk, September 18, 2026)
Ethereum's Glamsterdam hard fork activated on the Sepolia testnet on October 6, 2026, at 13:53:36 UTC, tripling the block gas limit from 60 million to 200 million. Early test blocks processed between 52 million and 92 million gas — 26% to 46% of the new ceiling — while 99.97% of eligible stake participated in finality voting. The upgrade introduces two protocol-level architectural changes: Enshrined Proposer-Builder Separation (ePBS) via EIP-7732 and Block-Level Access Lists (BALs) via EIP-7928. Together, they represent Ethereum's most consequential consensus-layer restructuring since the Merge.
The Hoodi testnet is scheduled for activation on October 26, 2026. Mainnet remains unconfirmed but is targeted for Q4 2026, with November 4 cited as a conditional date. If successfully deployed, the upgrade aims to push Layer 1 throughput toward 10,000 transactions per second, cut base-layer fees by an estimated 78.6%, and reduce MEV extraction by up to 70% — while attempting to preserve solo-validator economics.
The Glamsterdam upgrade activated on Sepolia at slot height corresponding to October 6, 2026, 13:53:36 UTC. CoinDesk analyzed more than 25 consecutive post-upgrade blocks, each consuming between 52 million and 92 million gas — a range of 26% to 46% of the new 200 million allowance. A six-minute voting period produced all scheduled blocks, with 99.97% of eligible stake contributing to finality, according to on-chain telemetry.
These numbers demonstrate that the testnet infrastructure can produce and finalize blocks under the expanded gas ceiling. However, Sepolia blocks used less than half their capacity, making the result a proof of headroom rather than evidence that the network can sustain near-200-million-gas blocks under full production load. The distinction matters: real mainnet conditions — with thousands of independent validators, heterogeneous hardware, and adversarial MEV searchers — will stress the system in ways a controlled testnet cannot replicate.
The Ethereum Foundation's testnet announcement names four headline EIPs — 7732, 7928, 8037, and 8038 — alongside more than a dozen smaller proposals, all bundled under meta EIP-7773.
EIP-7732 integrates proposer-builder separation directly into Ethereum's consensus layer. Currently, 80% to 90% of Ethereum blocks are produced using external relays such as MEV-Boost, an off-protocol system built by Flashbots. Under ePBS, the proposer commits to a block header, and a separate builder constructs the execution payload. No external relay is required.
The implications are structural. According to research cited by Thirdweb and CryptoAPIs, ePBS is projected to reduce MEV extraction by up to 70%. The mechanism increases the data propagation window from approximately two seconds to roughly nine seconds, providing critical headroom for larger payloads and parallelized execution without choking block dissemination.
By enshrining the builder market in protocol, Ethereum removes the censorship-by-relay risk that has been a persistent concern. Under the current MEV-Boost architecture, relays can selectively exclude transactions. ePBS shifts enforcement to the protocol itself, where builders must satisfy consensus rules or forfeit their bid.
For validators, the change is double-edged. Professional block builders will continue to earn the bulk of MEV revenue, but the extraction pathway becomes transparent and auditable. Solo validators retain fair access to proposer slots without needing to trust third-party relay operators.
EIP-7928 introduces Block-Level Access Lists (BALs), which record every account and storage location accessed during a block, along with relevant post-transaction state changes. This metadata enables a fundamental shift in how Ethereum processes transactions: parallel execution across multiple CPU cores during block validation.
Without BALs, validators must execute transactions sequentially to avoid state conflicts. With them, the execution engine can identify non-overlapping state accesses and process them simultaneously. According to TokenToolHub's technical breakdown, this parallelism is what makes a 200 million gas limit technically feasible without mandating proportional hardware upgrades.
BALs also enable executionless state reconstruction for light clients, reducing the barrier for lightweight network participation. The combination of BALs and the ePBS propagation window creates the technical foundation for sustained higher throughput.
The practical constraint is disk I/O. Parallel state reads generate concurrent disk access patterns that differ from sequential execution. While the upgrade is designed to keep hardware demands within reach of solo validators, the Ethereum Foundation has acknowledged that some increase in storage performance requirements is likely.
The economic stakes of a 200 million gas limit are significant. Research from CoinLaw's 2026 gas fee analysis indicates that every five-million-unit increase in the gas limit correlates with a 59.2% decline in median transaction fees. If the 200 million maximum were fully utilized — implying a 100 million target under Ethereum's current elasticity model — the fee reduction on Layer 1 would be substantial.
Separate analysis projects that gas repricing changes bundled with Glamsterdam could cut L1 fees by approximately 78.6%.
The supply-side consequences are already visible. As of October 9, 2026, Ethereum has net-issued approximately 778,413 ETH year-to-date. Total issuance stands at roughly 796,623 ETH, against burns of just 16,525 ETH from execution and blob fees. Burns offset only 2.07% of new issuance, leaving the supply approximately 0.64% higher than at the start of the year, according to CryptoSlate.
This trend predates Glamsterdam. The Dencun upgrade (March 2024) slashed Layer 2 fees by over 90%, routing most transaction activity through cheap blob data lanes rather than the base layer. Median L2 fees on Arbitrum, Optimism, Base, and Linea dropped from the $0.20–$1.00 range to $0.01–$0.10.
A conditional model estimates that Ethereum would need to burn approximately 2,992 ETH per day to offset current issuance. With daily burns running well below that threshold, the "ultrasound money" narrative — predicated on deflationary supply mechanics — remains structurally challenged. Glamsterdam's fee reductions could depress burn rates further, widening the gap between issuance and destruction.
The Sepolia data demonstrates block production under the expanded limit but does not confirm that validators can sustain near-200-million-gas blocks on mainnet. According to TradingView's analysis of the test results, the Glamsterdam test "leaves validator strain unresolved."
The concern is hardware economics. Validating 200 million gas blocks without BALs would require hardware upgrades beyond what most solo validators or small operators could afford, according to Chainlabo's validator guide. With BALs, the upgrade is designed to keep requirements manageable, but the parallel disk-read pattern has not been stress-tested at mainnet scale.
The ePBS component offers a partial counterweight. By separating block building from block proposing at the protocol level, ePBS ensures that solo validators need not compete on execution optimization — only on timely attestation and proposal. According to Everstake's analysis, even smaller home-staked nodes can compete fairly with professional builders under the new architecture.
The unresolved question is whether the combined effect — higher gas, parallel execution, and protocol-level builder separation — will shift the validator set's composition. If hardware demands rise enough to push out marginal solo operators, the network's decentralization properties could erode incrementally, even as the protocol's design intent is to preserve them.
Before the Sepolia activation, Ethereum developers issued a public warning about builder abuse vulnerabilities specific to the testnet environment. As reported by CoinDesk on September 18, 2026, a developer noted that disposable builder identities could be used to submit oversized bids and then withhold promised execution payloads after winning.
The attack vector works as follows: an attacker creates a builder identity using free test ETH, wins a block auction with an inflated bid, then refuses to deliver the execution payload. Clients can identify and reject the offending builder, but the attacker returns under a new identity and repeats. On mainnet, where real ETH collateral is at stake, this attack becomes economically prohibitive. On testnets, it is essentially free.
CryptoDaily reported that developers set a September 29 deadline for client teams to release Sepolia-ready software, allowing seven review days before the October 6 activation. The Ethereum Foundation confirmed that this testnet-specific risk does not represent a direct threat to mainnet funds.
| Milestone | Date | Status | |-----------|------|--------| | Sepolia activation | October 6, 2026 | Complete | | Sepolia analysis | October 8, 2026 | Complete | | Hoodi activation | October 26, 2026 | Scheduled | | Mainnet activation | Q4 2026 (Nov. 4 conditional) | Unconfirmed |
The Hoodi testnet date was confirmed during the Ethereum Core Developers Meeting (ACDE #247) held on October 8, 2026. The mainnet date remains contingent on Hoodi results and subsequent developer consensus. The Ethereum.org roadmap marks Q4 2026 as the target window but flags it as unconfirmed.
If mainnet activates in November, Glamsterdam would ship roughly 14 months after the Pectra upgrade (May 2025), maintaining Ethereum's approximate cadence of one major fork per year.
Glamsterdam is a structural upgrade, not a parameter tweak. By enshrining proposer-builder separation and enabling parallel transaction execution at the protocol level, Ethereum is rebuilding the machinery through which blocks are produced, validated, and finalized. The Sepolia results are encouraging but incomplete — a proof of concept, not a production certification.
The economic tension is clear. Glamsterdam's fee reductions benefit users and applications but further erode the fee-burn mechanism that underpins ETH's scarcity narrative. With supply growing at 0.64% year-to-date and burns covering just 2% of new issuance, the upgrade prioritizes throughput and usability over monetary properties.
Whether this trade-off strengthens or weakens Ethereum's competitive position depends on execution. A successful mainnet deployment would give Ethereum L1 capacity that approaches current L2 throughput levels, potentially recapturing economic activity that migrated off the base layer. A botched rollout — validator strain, builder exploits, or consensus bugs at scale — could delay mainnet beyond Q4 and hand competitive ground to faster-moving chains.
The data from Hoodi, expected October 26, will determine whether November mainnet is realistic or aspirational.