The Glamsterdam upgrade test network continued producing blocks normally on 17 Sep 2026 after adopting new block-production rules, according to reporting by CoinDesk’s Joon Hyoung Lee and republished by Bloomingbit. The rehearsal delivered clean results for client teams and set the stage for a possible deployment on Ethereum’s Sepolia testnet as soon as October 6.
Glamsterdam upgrade clears its rehearsal round
Developers ran the latest Glamsterdam upgrade rehearsal on a devnet on 17 Sep 2026 and reported uninterrupted block production throughout the exercise, according to CoinDesk’s coverage cited by Bloomingbit. About 60 minutes after launch, the Glamsterdam testnet raised its per-block gas limit from 60 million to 200 million, more than tripling available execution capacity inside every block. That kind of step-change in capacity is unusually large for a single testnet session and signals confidence from core developers that the underlying engine can carry the load without compromising consensus stability.
Nethermind posts a clean scorecard on access lists
Client team Nethermind ran the headline performance test on Block-Level Access Lists, the centerpiece feature shipped inside Glamsterdam, CoinDesk reported. The suite passed all 2,302 tests in the benchmark and processed 570.7 billion gas over a span of 3 minutes and 15 seconds, equivalent to roughly 2.9 billion gas per second. The combination of full pass-rate and sustained throughput is the kind of result engineering leads look for when green-lighting a fork for public testnets, and it gives client implementers a baseline they can replicate across Geth, Besu, Reth and Erigon.
Capacity target now sits at a 200 million gas floor
The 200 million gas figure is not a one-off stress number, CoinDesk noted. Repricings baked into Glamsterdam are explicitly designed to push the network toward a 200M floor, so what the testnet demonstrated on day one is meant to be the floor the protocol maintains once the fork activates on Sepolia. Bringing the floor up alongside Block-Level Access Lists is intended to give rollups and application developers predictable headroom rather than the volatile peaks-and-valleys that have characterized mainnet gas markets in recent cycles.
Sepolia deployment targeted for October 6
Developers are weighing a deployment of Glamsterdam to the Sepolia testnet on October 6, CoinDesk reported via Bloomingbit, with the detailed schedule set to be locked in at an official meeting rather than announced ad hoc. Sepolia is the long-running public proof-of-stake testnet that most wallet, custody and staking teams use for pre-production sign-off, so a coordinated go-live date matters for anyone shipping validator or rollup code in the back half of 2026. Until that official meeting closes, October 6 should be read as a target, not a guarantee.
Two forks merged into one ambitious package
Glamsterdam is the combined shipping vehicle for two previously separate forks, Gloas and Amsterdam, that Ethereum researchers had been planning in parallel. Folding them together lets the protocol land a large architectural shift in a single coordinated activation rather than dragging the ecosystem through two consecutive upgrades. CoinDesk described the combined scope as one of the more consequential forks Ethereum has attempted since the Merge in September 2022.
ePBS, contract sizes and forward-compatible consensus
The headline features inside Glamsterdam include Enshrined Proposer-Builder Separation, Block-Level Access Lists, a stack of gas repricings aimed at that 200 million floor, larger contract limits (24KiB to 64KiB deployed and 48KiB to 128KiB initcode), and forward-compatible consensus data structures, CoinDesk reported. ePBS in particular is the change most analysts have been watching, because it moves proposer-builder separation from an out-of-protocol marketplace such as MEV-Boost into the consensus layer itself. Larger contract caps, in turn, reduce bytecode-splitting gymnastics for developers shipping complex Solidity and Vyper applications, while forward-compatible consensus structures are meant to spare Ethereum another hard fork when future scaling changes land.
Why node costs are not expected to spike
The architectural rewrite is explicitly designed to expand capacity without excessively raising node operating costs, CoinDesk emphasized. That matters because every previous step-up in gas limits has been accompanied by fears that home stakers and solo validators would be priced off the network. The pitch with Glamsterdam is that Block-Level Access Lists and the 200M repricings deliver more useful work per block while keeping witness verification, state access and disk growth within reach for consumer-grade hardware. Whether that promise holds on mainnet will be one of the bigger questions for Ethereum node operators once Sepolia and then Holesky have absorbed weeks of stress testing.
What to watch next on the Glamsterdam roadmap
Three dates now anchor the near-term Glamsterdam upgrade roadmap: the official scheduling meeting that will confirm or shift October 6, the Sepolia activation itself, and a subsequent Holesky testnet phase that historically precedes mainnet. Stakeholders including staking pools, rollup sequencers, wallet vendors and block-builder operators will be tracking each milestone for client release notes and validator guidance. If the September 17 devnet results hold up under public testnet conditions, the path toward a 2026 mainnet Glamsterdam upgrade looks concrete for the first time since the Gloas-and-Amsterdam merger was announced.
The clean run on day one of the devnet, the triple-digit gas limit bump and a confirmed 2,302-of-2,302 score on Block-Level Access Lists together give the Glamsterdam upgrade its strongest signal yet that the long-gestating combination of Gloas and Amsterdam is finally approaching liftoff on public testnets, with Sepolia activation penciled in for October 6 and a final schedule expected to be ratified at an official developer meeting in the coming days.
Source: CoinDesk via Bloomingbit, Joon Hyoung Lee, 17 Sep 2026