Ethereum's Glamsterdam Devnet Extends the Execution-Payload Window From Two Seconds to About Nine

Ethereum's next scheduled upgrade, Glamsterdam, now has a public devnet where validators and infrastructure operators can test its changes before they reach the established testnets and mainnet. The headline mechanical change is timing. Today a builder has roughly two seconds to propagate and process an execution payload inside the slot. Under enshrined proposer-builder separation, the consensus block and the execution payload are handled as separate steps, and the propagation window widens to about nine seconds.

Those seconds are the constraint that has kept Ethereum's gas limit low. Validators currently coordinate their gas votes around the 60 million range. A wider window gives nodes more time to receive and verify larger blocks and more blob data, which is what makes a much higher gas ceiling achievable without slowing block propagation. The fork itself does not set that ceiling. Validators still raise it by signaling, and only as real nodes demonstrate they can keep up.

Slot timing is also central to Ethereum's known consensus attacks. A contribution to the research forum on timing the head in Ethereum proof-of-stake sets out the mechanism plainly: many reorganization attacks work by manipulating what validators believe the current head of the chain is, splitting attestation weight across competing branches. Re-cutting a slot into separately timed consensus and execution phases changes the surface those attacks operate on, which is exactly what a devnet exists to stress before real value depends on it.

There is a commercial consequence as well. Ethereum holds $47.36 billion of the $84.86 billion in total decentralized-finance value tracked by DefiLlama, and the layer-2 networks built on it secure $42.30 billion in total, with Base at $12.12 billion and Arbitrum One at $11.07 billion. Rollups exist partly because base-layer blockspace is scarce and expensive. If Ethereum raises throughput and cuts settlement costs by repricing its own protocol, the fee gap that justifies moving activity off the base layer narrows, and each rollup must compete on something other than cost.