News Ethereum and other coins Blockchain and technology
Ethereum’s Glamsterdam test: a 200 million gas target on Sepolia
A new validator software release sets a higher gas target on the Sepolia testnet at the scheduled Glamsterdam upgrade. The change concerns a network-capacity test; it is not an announcement of mainnet activation or fee reductions already achieved.
Checked on 6 October 2026.
The Prysm team released version 7.2.1 on 5 October, with a change directly relevant to Ethereum's next test: validators using it will set a default gas limit of 200 million from the Gloas activation point on Sepolia. According to the developers' release announcement (Opens in a new tab), that point was scheduled for 6 October 2026 at 13:53:36 UTC, or 16:53:36 in Israel.
Gloas is the consensus component of Glamsterdam, the upgrade changing the division of work in block building and validation. The client update illustrates the challenge: testing greater capacity requires more than agreeing on protocol rules. Operators' software and settings must also match them.
What 200 million gas means
Gas measures computational work on the network, and the gas limit determines how much work a block can contain. It is not the number of transactions in a block, because different transactions consume different amounts of resources.
The new Prysm version defines a schedule applying the higher default at the upgrade point. The developers note that operators can set a different limit. The 200 million figure therefore describes a client setting for the Sepolia test, not proof that every block reaches that limit or that mainnet has adopted it.
The same release also fixes the process for producing the first block after the upgrade and changes block-builder settings. Capacity testing depends not only on the selected limit but also on network components working together.
ePBS: separating roles within the protocol
One of Glamsterdam's central changes is ePBS, a mechanism described in EIP-7732. It separates the block proposer from the builder of its execution payload and incorporates their commitment and payment exchange into the network's rules.
The separation also allows validation to be broken into stages: agreement on a block need not wait for all execution work to finish within the same tight time window. The aim is to allow more time for processing data and reduce reliance on external intermediaries for exchanging payloads and payments. This changes the operating mechanism; it does not promise a particular outcome for ETH's price.
Access lists that allow parallel work
Another component, BALs, is described in EIP-7928. These are block-level lists recording which accounts and storage locations were accessed during execution, alongside the resulting changes.
The lists give node software an explicit picture of the data involved. This allows parallel disk reads and more efficient transaction validation and calculation of network state. Increasing capacity therefore involves more than raising a numerical ceiling: it requires a more efficient way to check the work included in each block.
What changes for ETH holders
The Ethereum Foundation's scheduling announcement set a date for Sepolia only. Activation dates for Hoodi and mainnet remained unconfirmed. The announcement also makes clear that ETH holders and mainnet users do not need to act because of the testnet upgrade.
Sepolia operators must use supporting clients at both layers, and application developers need to test compatibility with gas and interface changes. For readers following Ethereum, the test is whether the changes work reliably and efficiently before a decision is made to apply them to the main network. A scheduled date, software version and capacity target do not, by themselves, confirm that the test has succeeded.