Ethereum: Glamsterdam Clears Its First Major Test on Sepolia

Ethereum has just activated Glamsterdam on the Sepolia testnet. The crypto upgrade was launched on October 6 at 13:53 UTC, at epoch 353,024. Several important changes are now being tested on a public network: new organization of block construction, preparation for parallel processing, and revision of certain gas costs. The mainnet will still have to wait.

Glamsterdam finally passes on Sepolia

Ethereum had set October 6 as the date for the public test of Glamsterdam. The activation took place at slot 11,296,768, i.e., at 13:53:36 UTC. Sepolia represents a very different step from the private networks used during development.

The testnet gathers more operators, software, and different configurations. Teams can thus verify that execution and consensus clients all correctly apply the new rules.

Node operators had to install compatible versions before the fork. This notably concerns software like Geth, Besu, and Reth for execution, as well as Prysm, Teku, or Lodestar for consensus.

ETH holders have nothing to do. Sepolia ethers are only test tokens and Glamsterdam is not yet active on the Ethereum mainnet.

Nevertheless, the move to Sepolia was not guaranteed a few weeks earlier.

The devnets had encountered several finality issues and a bug related to the new gas pricing. Ethereum still had to fix several problems before exposing Glamsterdam to a public network. This phase is now behind it. The work now changes terrain.

ePBS modifies Ethereum block construction

Glamsterdam contains a particularly important modification: ePBS, for enshrined proposer-builder separation. Ethereum already uses a separation between the validator who proposes a block and the specialized actors who build its content.

The builders notably seek to organize transactions efficiently before proposing their blocks to the validators. Part of this architecture currently depends on software and infrastructures located outside the protocol itself.

EIP-7732 aims to integrate this organization directly into Ethereum’s rules. The builder prepares the block. The proposer publishes the corresponding commitment. The protocol then organizes the content revelation, its validation, and the payment.

The change may seem very technical, but it directly affects how Ethereum produces its blocks. One of the objectives is to give validators more time to verify execution. With heavier blocks, every millisecond available becomes more important.

ePBS must also reduce Ethereum’s dependence on certain external infrastructures currently used to coordinate builders and validators.

Block construction thus becomes more native to the network. This project is especially important in Ethereum’s crypto strategy: the protocol wants to significantly increase its capacity without asking validators to simply process more data in the same amount of time. Glamsterdam therefore begins by reorganizing the mechanics before increasing the load.

Block-Level Access Lists prepare parallel processing

Another major novelty concerns the Block-Level Access Lists, or BAL. When an Ethereum block contains many transactions, these interact with different accounts, smart contracts, and storage locations.

Currently, clients have to discover some of these accesses during execution. EIP-7928 adds a list indicating the state elements used by a block and certain modifications made to them.

This information opens the door to more parallel work. If two operations do not use the same data, clients can prepare or process part of their execution simultaneously rather than always waiting for each previous operation to complete.

Ethereum does not instantly become a fully parallel blockchain.

BAL mainly provide one of the building blocks needed to move in this direction. They can also facilitate disk data reads and certain calculations performed post-execution. The aim is directly related to the time needed to process a block.

The more Ethereum wants to accept transactions, the more this issue becomes important. Simply increasing block sizes without speeding up their processing could make nodes more expensive to operate and complicate validation.

This is where Glamsterdam joins the broader network goal. Developers now consider that a capacity of 200 million gas after Glamsterdam becomes technically feasible. Ethereum currently runs around 60 million. The scale change would be substantial.

The gas cost also needs to evolve

More capacity creates another problem: not all Ethereum operations cost the same to the network. A transaction might require computation for a few moments without leaving much data behind. Another might create new information that nodes will need to keep permanently.

However, historical gas pricing does not always perfectly reflect this difference. Glamsterdam must correct part of this imbalance. EIP-8037 notably introduces a more precise way to measure the cost associated with new state creation.

The goal is to prevent throughput increase from blowing up the size of data each node must store. EIP-8038 also revises the cost of several state access operations. As a result, some operations might cost more gas after the upgrade.

This is intentional. Ethereum does not only seek to reduce fees per transaction. It mainly wants each operation to pay a cost closer to the resources it actually consumes. Some applications will have to check their code. Tools that use hardcoded gas limits may encounter problems. Fee estimators, wallets, and development infrastructures have also been invited to test the new rules before mainnet rollout.

A simple transfer to an existing account will not necessarily be affected the same way as an operation creating new permanent data. Scaling thus also involves pricing. More space, but not free for all operations.

Hoodi must now open the path to the mainnet

Sepolia is not the final destination. Ethereum still has to deploy Glamsterdam on Hoodi, its other major test environment notably used by validators and staking infrastructures.

No definitive date has been announced yet. The mainnet doesn’t have one either. Developers want to first observe how Sepolia reacts to the new rules: finality, client compatibility, ePBS behavior, block creation, BAL functioning, and new gas rules.

A major problem could still delay the schedule. The stakes are high. Ethereum secures hundreds of billions of dollars in assets, stablecoins, DeFi protocols, and tokenized products. An unnoticed error on a test network could become much more costly if it happens on the mainnet.

Testing is precisely to avoid this. After Glamsterdam, Ethereum is already working on Hegotá, the next big upgrade planned for 2027. Several proposals are still under study, notably around privacy and censorship resistance.

But the immediate project remains Glamsterdam. Sepolia is now active. Hoodi must follow. Only then will the main network come. If everything works correctly, Ethereum will have a much stronger technical base to push its Layer 1 beyond current limits.

The 200 million gas will not be activated automatically with Glamsterdam. Validators will have to gradually raise their parameters and the network will have to demonstrate it can absorb the load. This is exactly what the next tests are meant to measure. Glamsterdam does not yet accelerate Ethereum for all its users. But the first major public test of this new architecture is now underway.