Ethereum's Sepolia testnet has activated the Glamsterdam upgrade, a change that includes an optional 200 million gas limit. The upgrade is a testbed signal: it points to how much more throughput Ethereum might eventually handle, while also surfacing concerns about what that extra capacity does to decentralization and node performance.
What Glamsterdam actually changes on Sepolia
The headline number is the 200 million gas limit, which is optional rather than mandatory. That distinction matters. Node operators and developers on Sepolia can choose whether to adopt the higher ceiling, which means the testnet won't produce a single uniform picture of how the network behaves under load. Some participants will push toward the maximum, others will stay where they are, and the data coming out of the testnet will reflect that mixed environment.
Sepolia exists for exactly this kind of experiment. It's a proving ground before anything touches mainnet, and Glamsterdam's presence there is the first real look at how the broader Ethereum ecosystem reacts to a substantially higher gas ceiling.
The decentralization question nobody's settled
Higher gas limits sound like an easy win. More room per block, more transactions, cheaper activity. The trade-off is that bigger blocks take more resources to process, and not every node operator runs the same hardware. Raise the ceiling too far and the set of people who can realistically validate the chain shrinks. That's the decentralization concern, and it's the one that tends to get waved away in capacity discussions until it becomes a problem.
On Sepolia, that tension is now testable rather than theoretical. Node performance under the optional 200 million gas limit will be the thing developers watch. If weaker machines start falling behind, that's a signal worth paying attention to before any mainnet conversation gets serious.
Glamsterdam's Sepolia activation is a hint, not a commitment. Ethereum's capacity roadmap has been debated for years, and every step toward higher throughput runs into the same wall: the network's strength comes from how many independent operators can keep up. A testnet run doesn't answer that. It just gives the ecosystem real numbers to argue over.
What happens next is straightforward. Developers will watch how Sepolia performs with the optional limit in place, and the results will feed into whether the same approach gets proposed for mainnet. There's no hard timeline attached to that, and the concerns around node performance and decentralization won't resolve themselves just because a testnet activated cleanly. The real question is whether the data from Sepolia makes the case for higher capacity or undercuts it. That's still open.
