Ethereum Glamsterdam Nears 200M Gas Limit Test

Highlights

Ethereum's Glamsterdam upgrade completed its Devnet-11 rehearsal this week, confirming blocks under new block-building rules
The test network raised its block gas limit toward 200 million, more than triple today's 60 million cap
The Nethermind client passed all 2,302 performance tests, processing 570.7 billion gas in three minutes and 15 seconds
Developers are targeting an October 6 deployment on the public Sepolia testnet, still pending formal confirmation

Ethereum's Glamsterdam hard fork reached its final devnet stage this week, with core developers confirming blocks successfully under the upgrade's new block-building rules during the Devnet-11 rehearsal. The test locked in the bundle of Ethereum Improvement Proposals — tracked collectively under Glamsterdam's Meta EIP-7773 — that developers now expect to carry through public testnets toward a mainnet activation targeted for the second half of 2026.

The headline number to come out of the test is capacity. Devnet-11 pushed the block gas limit toward 200 million, more than triple Ethereum mainnet's current 60 million cap, as developers prepare for a proposed public test on the Sepolia testnet on October 6. That date and the 200 million figure both remain provisional: CoinDesk reported that the higher limit is still a test setting rather than a mainnet commitment, and that it has yet to undergo the kind of adversarial stress-testing — deliberate attempts to break or exploit the larger blocks — that would typically precede a production rollout. Devnet-11 itself was designed as a controlled rehearsal rather than an adversarial test, meaning the real stress test still lies ahead.

Two EIPs are doing the structural heavy lifting

Of the ten EIPs bundled into Glamsterdam, two stand out as the core architectural changes. EIP-7732 enshrines Proposer-Builder Separation directly into the protocol, formalizing a division of labor between the entities that propose blocks and those that build them — a split already used informally today through third-party relays, but not yet a native part of Ethereum's consensus rules. EIP-7928 introduces Block-Level Access Lists, which let validators identify in advance which parts of the state a given block's transactions will touch, enabling nodes to fetch data and verify unrelated transactions in parallel rather than processing everything sequentially. That parallelization is a large part of what makes a much larger block gas limit practical without proportionally slowing down validators.

Handling bigger blocks also means giving the network more time to spread the added data around before the next block arrives. Glamsterdam extends the window for block data propagation from roughly two seconds today to about nine seconds, giving validators more breathing room to receive and check larger blocks before they need to vote on them. On the performance side, the Nethermind execution client — one of several client implementations that must independently handle Ethereum's rules — passed all 2,302 performance tests run against it during the devnet, processing 570.7 billion gas in three minutes and 15 seconds. The capacity push is running on a separate track from Ethereum's other major infrastructure priority this year: recent US quantum grants aimed at helping Ethereum and Bitcoin race post-quantum security deadlines, a reminder that scaling and long-term cryptographic security are being worked on as parallel, not sequential, priorities.

None of this locks in a final mainnet gas limit; that decision still runs through Ethereum's usual client-team and community coordination process, and a 200 million ceiling that performs well in a scripted devnet rehearsal is a different bar than one that survives real adversarial conditions on a public testnet. But clearing Devnet-11 without failures gives the Glamsterdam roadmap a concrete data point to build the October Sepolia push around.

The move to public testnets typically brings a different class of stress than a controlled devnet rehearsal, since Sepolia is open to outside validators and researchers actively looking for ways to break new protocol rules rather than a curated set of client teams running scripted scenarios. If the 200 million gas limit and its supporting infrastructure — the block-level access lists, the extended propagation window — hold up under that kind of open adversarial testing, it would mark one of the more significant capacity increases in Ethereum's history since the merge, though developers have been careful to stress that a mainnet figure could still land below the devnet's test ceiling depending on what Sepolia turns up. Ethereum co-founder Vitalik Buterin has separately put a 60% chance on cryptographic tools like SNARKs and FHE hitting sub-10x overhead within a similar timeframe, underscoring how much of Ethereum's roadmap right now is running on parallel, still-unproven technical bets.