Ethereums next major upgrade is no longer only a roadmap document.
Glamsterdam activated on the Sepolia public testnet on October 6, 2026. Early upgraded epochs finalized, and the testnet began a gradual increase from a 60 million gas limit toward a 200 million client target.
The number is attention-grabbing.
The architectural changes underneath it matter more.
Glamsterdam combines changes to Ethereums execution and consensus layers. Its two headline features are enshrined proposer-builder separation, or ePBS, and block-level access lists, or BALs.
Together, they are intended to let Ethereum process and validate more execution without making the validator workload scale in the same naïve way.
That is why Glamsterdam should not be reduced to “Ethereum raises the gas limit.”
The upgrade is testing a different way to build and validate large blocks.
A 200M Gas Target Does Not Mean Mainnet Has 3.3x Capacity Today
Before the Sepolia fork, major client configurations were centered around a 60 million gas limit.
The Glamsterdam test pushes the test network toward a 200 million target.
That does not mean Ethereum mainnet has already increased capacity by more than three times.
Sepolia is a public testing environment.
Hoodi and mainnet activation dates had not been announced at the time of writing.
The purpose of the higher limit is to expose clients, validators, builders and applications to the operating conditions Ethereum may eventually want to support.
A large block is only useful if the network can propagate, execute and verify it safely.
The test therefore needs to answer more than “did the block fit？”
It needs to answer:
- did nodes keep up;
- did finality remain stable;
- did builders reveal payloads on time;
- did state access remain manageable;
- did gas estimation behave correctly;
- did clients remain interoperable.
Capacity is an end-to-end property.
ePBS Moves the Builder Relationship Into Ethereum Consensus
Ethereum already has a separation between block proposers and specialized block builders.
Much of that market evolved through off-protocol infrastructure such as MEV-Boost.
Glamsterdams EIP-7732 brings proposer-builder separation into the consensus protocol itself.
A proposer includes a commitment from a builder.
The builder then reveals the execution payload.
A payload timeliness committee helps determine whether the payload and associated blob data arrived on time.
The protocol also handles the payment relationship between builder and proposer.
The long-term objective is to reduce reliance on trusted middleware for a function that has become economically central to Ethereum block production.
This is significant because builder infrastructure is not a niche component.
It sits directly in the transaction-ordering and MEV supply chain.
Enshrining the relationship changes who Ethereum depends on to coordinate that market.
Why Separating Consensus and Execution Validation Helps Scaling
A validator has a limited amount of time inside each slot.
If it needs to receive, execute and fully validate an increasingly large payload before performing all other consensus duties, higher throughput eventually collides with slot timing.
ePBS separates parts of that process.
Consensus can progress around a builder commitment while execution-payload validation has a more explicit path.
This is not a free increase in throughput.
Builders and validators still need to process the data.
The design changes the timing budget.
That timing budget is one reason Glamsterdam can contemplate larger blocks without simply assuming every validator becomes several times faster.
Block-Level Access Lists Make Parallel Validation More Practical
EIP-7928 adds enforced block-level access lists.
The lists record the accounts and storage locations touched during a block, along with post-transaction state information.
Why does that matter？
Ethereum transactions frequently access shared state.
A client that does not know which pieces of state will be needed may have to fetch data serially during execution.
If the block already describes its state-access pattern, clients can prepare data and perform more work in parallel.
That can improve:
- disk access;
- parallel transaction validation;
- state-root computation;
- state synchronization.
BALs therefore attack a different scaling bottleneck from block size.
They try to make verification itself more parallelizable.
This fits Ethereums broader post-Fusaka direction: more throughput is useful only if independent nodes can still verify the chain.
Gas Repricing Is Part of the Same Scaling Strategy
Glamsterdam also changes how Ethereum prices computational and state resources.
EIP-8037 increases and separately meters state-creation costs.
EIP-8038 changes state-access costs.
Other included proposals adjust calldata, intrinsic transaction gas, access lists and block gas accounting.
This is not bookkeeping trivia.
If some operations consume more long-term node resources than their gas price reflects, raising the gas limit amplifies the mismatch.
A sustainable scaling plan therefore needs both:
more total capacity and more accurate pricing of expensive resources.
Application developers need to test assumptions around hardcoded gas limits, gas stipends and transaction estimation.
A contract that worked economically under one gas schedule may behave differently after repricing.
The Last-Minute Prysm Fix Shows Why Testnets Matter
One of Ethereums major consensus clients, Prysm, shipped version 7.2.1 shortly before activation.
The previous version supported the fork but did not automatically use the new 200 million gas preference without additional operator configuration.
The updated release changed the default for the test.
This is exactly the kind of issue a public testnet rollout is designed to expose.
It was not a consensus failure.
It was an operational configuration mismatch that could have caused validators to propose materially smaller blocks than the intended capacity test.
Large protocol upgrades fail as often through tooling assumptions and operational defaults as through elegant specification errors.
Sepolia exists to find those problems before mainnet.
Why It Matters
Ethereums scaling debate is often presented as a conflict between Layer 1 and Layer 2.
Glamsterdam shows that Ethereum is still investing aggressively in the L1 itself.
The objective is not to replace rollups.
It is to make the settlement and execution base more capable while preserving verifiability.
ePBS addresses the block-building pipeline.
BALs address parallel validation.
Gas repricing addresses state and execution economics.
The higher gas target tests how far those changes can push real capacity.
Taken together, the upgrade is better understood as a verification architecture upgrade than a simple throughput upgrade.
Risks and Counterarguments
Glamsterdam is live on Sepolia, not mainnet.
A successful testnet activation does not prove the same capacity is safe under mainnet economic load.
ePBS introduces new protocol roles and timing logic around builders.
Block-level access lists increase block metadata and implementation complexity.
Higher gas limits can increase hardware and bandwidth demands if efficiency gains do not compensate.
Gas repricing can break assumptions in existing applications.
Mainnet timing remains unknown.
Ethereum also needs client diversity: a design that works only under one implementation is not robust decentralization.
What to Watch Next
The next useful signals are operational rather than promotional:
- Sepolia finality and missed-block rates;
- how quickly the gas limit reaches and sustains higher levels;
- builder reveal failures;
- client-specific issues;
- block propagation;
- state growth;
- application gas regressions;
- Hoodi activation;
- the eventual mainnet schedule.
The key question is not whether Sepolia can produce a 200 million gas block.
It is whether Ethereum can make much larger blocks routine without making independent verification materially less practical.
FAQ
Is Glamsterdam live on Ethereum mainnet？
No. It activated on Sepolia on October 6, 2026. Hoodi and mainnet dates remain to be announced.
What is ePBS？
Enshrined proposer-builder separation moves the proposer-builder relationship into Ethereums consensus protocol instead of relying entirely on external middleware.
What are block-level access lists？
They describe the state locations a block touches, helping clients fetch and validate state more efficiently and in parallel.
Did Ethereum raise mainnets gas limit to 200 million？
No. Sepolia is testing a move toward a 200 million client target.
Why is the upgrade important？
It tests whether Ethereum can increase L1 execution capacity by redesigning block production, validation and gas economics rather than simply asking validators to do more work.
