Ethereum's next planned network upgrade, codenamed Glamsterdam, has cleared a key capacity rehearsal on its devnet, confirming that the network can continue producing blocks under the upgrade's new block-building rules ahead of a proposed Sepolia testnet deployment targeted for October 6, 2026.

What the Capacity Rehearsal Validated

A capacity rehearsal is a pre-deployment stress test designed to prove that a proposed block-building rule set can sustain chain liveness and handle elevated gas throughput before an upgrade advances to a public testnet. Clearing the rehearsal signals that the devnet neither stalled nor forked under load, but it is not an upgrade activation, nor a final performance guarantee for mainnet conditions.

The most visible test parameter was the block gas limit, which was raised from 60 million to 200 million roughly an hour after the devnet went live. CoinDesk reported that the 200 million figure is a devnet test setting, not a commitment to that parameter on mainnet.

On the execution-client side, Nethermind, one of Ethereum's major client implementations, ran 2,302 performance tests, processing 570.7 billion gas in 3 minutes and 15 seconds in a standalone client benchmark. That figure measures the execution software's throughput in isolation, not end-to-end network speed across validators and proposers. The distinction matters because every node on the network must re-execute each block to verify it, so execution throughput sets a practical ceiling on how large blocks can grow.

Why the Rehearsal Matters for Glamsterdam

Glamsterdam's two headline EIPs are central to why capacity validation is consequential. EIP-7928, which introduces block-level access lists, reorganizes how transactions are processed and state is managed at the block level, while EIP-7732, known as ePBS (enshrined proposer-builder separation) — a change that writes the split between building blocks and proposing them into the protocol's own rules — extends the data-propagation window from roughly 2 seconds to roughly 9 seconds, giving block builders and validators more time to assemble and attest to larger, more complex blocks.

That longer propagation window is what makes higher gas limits operationally plausible without forcing node operators to upgrade hardware. If execution clients cannot process a larger block within the extended window, the block risks being orphaned — dropped from the canonical chain despite having been validly produced — creating MEV extraction pressure and the kind of consensus instability MEV bots can exploit. MEV, short for maximal extractable value, refers to profit captured by reordering, inserting, or censoring transactions within a block. The devnet run demonstrated that at least one major client can keep pace with elevated throughput under the new timing constraints.

For DeFi users and protocol teams, the rehearsal result is a readiness signal rather than a launch confirmation. A higher L1 gas limit would compress per-unit gas costs for complex interactions, which matters for protocols running multi-hop swaps, liquidation sequences, or large batch operations on Ethereum mainnet. None of that changes, however, until the upgrade clears further testing and formal scheduling.

What Comes Next Before Glamsterdam Advances

The official Ethereum roadmap lists the Sepolia fork as the next milestone, currently scheduled for October 6, 2026, though CoinDesk notes the date still requires formal confirmation from core developers. Sepolia is a public testnet, so deployment there will surface issues a single-operator devnet cannot, including cross-client compatibility among Nethermind, Geth, Besu, and Erigon under real validator set conditions. Because Ethereum relies on multiple independently developed clients to enforce the same rules, mismatches in how each implements Glamsterdam's new block-building rules are the kind of issue that emerges only once diverse operators with different hardware and configurations begin producing and verifying blocks together.

After Sepolia, the standard upgrade path runs through Holesky before any mainnet scheduling decision. Core developers will review client interoperability data, gas-limit parameter proposals, and any issues flagged at the governance level through the ethereum/pm issue tracker before setting mainnet fork block. Developers and node operators should track those channels, not devnet results, for production timelines.

The protocol's approach of separating execution-client benchmarks from full-network rehearsals mirrors how prior upgrades such as Dencun were staged. Confidential transaction infrastructure on Ethereum, such as Zama's private swap layer built on Morpho, will also need to verify compatibility with ePBS's modified proposer-builder separation before those integrations can rely on Glamsterdam's timing guarantees.

As the devnet result circulated, ETH was trading at $2,476.44, up 3.4% on the day, according to CoinGecko data, with the crypto Fear & Greed Index sitting at a neutral reading of 50.