Glamsterdam explained: what Ethereum's next upgrade changes and why it keeps slipping
ePBS, block-level access lists and a 200 million gas limit: a plain guide to Ethereum's next hard fork and the gas-pricing issue that halted its Hoodi launch.
ETHEREUMBLOCKCHAINANALYSIS
Daniel Hartley
10/27/20263 min read
Ethereum's next hard fork, Glamsterdam, was meant to be the big upgrade of 2026. It went live on the Sepolia test network on October 6, but on October 22 developers called off its planned October 26 activation on the larger Hoodi testnet after a security review raised concerns about a set of gas changes. Ethereum researcher Nixo has said the delay will more than likely push the mainnet launch into 2027. Here is what the upgrade does, why it matters and what caused the latest hold-up.
Like recent forks, Glamsterdam combines two upgrades: Amsterdam on Ethereum's execution layer, which processes transactions, and Gloas on the consensus layer, which coordinates validators. Its two flagship changes are enshrined proposer-builder separation, known as ePBS, and block-level access lists.
Enshrined proposer-builder separation
Today, most Ethereum blocks are not assembled by the validators who propose them. Specialized builders compete to construct the most profitable block, and validators pick the best bid through third-party software called MEV-Boost, which relies on trusted intermediaries known as relays. The system works, but it puts a small number of relays in a sensitive position: they see blocks before publication and could, in principle, censor transactions or go offline.
ePBS, specified in EIP-7732, moves this auction into the protocol itself. Builders commit to blocks and payments in a way the protocol can enforce, so validators no longer need to trust relays. A side benefit is timing. Separating the commitment from the delivery of a block's contents gives the network more time within each slot to propagate and verify data, and that extra slack is one of the things that makes higher throughput safer.
Block-level access lists
The second feature, defined in EIP-7928, requires each block to state in advance which accounts and storage slots its transactions touch and what the resulting values are. With that map in hand, Ethereum clients can load data in parallel and execute independent transactions at the same time rather than strictly one after another. For node operators, it means faster processing on the same hardware. For the network, it opens room to raise capacity. That capacity is the point. Developers paired Glamsterdam's test deployments with a jump in the block gas limit, effectively the size of each block, from 60 million to 200 million on Sepolia, with the same increase planned for Hoodi. A higher gas limit means more transactions per block on the base layer and, all else being equal, lower fees.
Why Hoodi was postponed
The upgrade also reprices certain operations so that gas costs better reflect the burden they place on nodes, particularly creating and accessing state. Two proposals, EIP-8037 and EIP-8038, raise the cost of those operations. According to Bankless, security firm ChainSecurity found that the new prices could change how some existing contracts behave. Writing to a fresh storage slot, for example, could cost around 110,000 gas instead of about 22,100, which can exceed fixed gas allowances hardcoded into older contracts.
The worry is an attack pattern called insufficient gas griefing, in which someone submits a call with too little gas so that a legitimate action fails and the failure is recorded as final. ChainSecurity said it had found 13 real-world cases and was contacting the affected projects. Developers decided they needed time to study the interaction and removed the planned Amsterdam and Gloas activations, along with the 200 million gas limit, from Hoodi.
What it means for users and builders
For ETH holders and ordinary users, nothing needs to be done; the delay affects test networks and client teams. For developers, the episode is a reminder that gas costs are part of a contract's environment, and that code which assumes fixed costs can break when the protocol changes. Projects with older contracts that forward fixed gas amounts should review them now rather than wait for a mainnet date.
For the broader roadmap, a slip in Glamsterdam has knock-on effects. The following fork, Hegotá, is being scoped in parallel, with FOCIL, a mechanism that lets several validators force transactions into blocks to strengthen censorship resistance, among the leading candidates. Vitalik Buterin has described Hegotá as likely the last conventional fork before a longer rebuild of the protocol, so delays at this stage push back a long chain of work.
The next milestones are a revised specification for the gas repricing, new client releases, a fresh Hoodi date and, possibly, a second fork on Sepolia if the specification changes. Until those arrive, any mainnet date for Glamsterdam is guesswork. What is clear is that Ethereum's core developers have chosen caution over schedule. For a network that secures hundreds of billions of dollars in value, that is the trade-off most users would want.