Ethereum’s scaling strategy is changing in emphasis. The network still relies on Layer 2 rollups for low-cost execution, but core developers now devote more work to expanding Layer 1 capacity. Glamsterdam, the next major upgrade, brings that change into clearer view through enshrined proposer-builder separation and block-level access lists. Ethereum targets Q4 2026 for mainnet activation, although developers have not confirmed a final date.
Glamsterdam Puts Layer 1 Capacity Back in Focus
For several years, Ethereum’s public scaling story centered heavily on rollups. Layer 2 networks execute transactions away from mainnet and use Ethereum for data availability, settlement, and security. That model still sits inside Ethereum’s roadmap. However, the Ethereum Foundation merged its separate “Scale L1” and “Scale Blobs” work into one Scale track in 2026, reflecting closer coordination between mainnet execution and rollup data capacity.
Glamsterdam turns that coordination into protocol changes. Ethereum’s roadmap says the upgrade will reorganize block creation, transaction processing, and database management to support higher Layer 1 throughput.
Core developers have also worked toward gas limits above 100 million, while an Ethereum Foundation update in May described 200 million gas as a credible post-Glamsterdam floor if the related optimizations perform as expected.
That target does not guarantee a fixed mainnet setting, but it shows how much more capacity developers now expect from L1.
ePBS Gives Ethereum More Room to Process Blocks
EIP-7732 defines enshrined proposer-builder separation and moves the relationship between block proposers and block builders into the Ethereum protocol. Today, many validators use external infrastructure to outsource block construction. ePBS creates an in-protocol builder role and separates the timing of consensus work from execution payload validation.
That separation gives validators more time to handle larger blocks and more data. Ethereum’s Glamsterdam documentation says ePBS can extend the critical propagation window from roughly two seconds to about nine seconds. The design also reduces reliance on trusted middleware between proposers and builders. These changes do not directly increase transaction throughput on their own, but they remove timing limits that constrain future gas and data increases.
Block-Level Access Lists Target Sequential Bottlenecks
EIP-7928 introduces block-level access lists to address another Layer 1 constraint. Ethereum nodes normally learn which accounts and storage locations a transaction touches during execution. That process makes parallel work difficult because clients cannot easily identify independent transactions before running them.
BALs give clients an upfront record of state locations that a block accesses, along with post-execution values. Clients can use that information for parallel disk reads, faster validation, parallel state-root work, and executionless state updates.
Ethereum also plans networking changes that let peers exchange these access lists during synchronization. Together, those changes prepare clients for more parallel processing without forcing every node to replay work in the same strictly sequential way.
Layer 2 Still Has a Defined Role
The stronger focus on Layer 1 does not remove Layer 2 from Ethereum’s scaling plan. Ethereum Foundation researchers now describe L1 as a permissionless settlement, liquidity, shared-state, and DeFi hub, while Layer 2 networks provide extra scale, customization, and specialized features. That framing differs from the earlier idea that L2s would carry nearly all future execution growth.
Glamsterdam therefore looks more like a rebalancing than a reversal. Ethereum continues to expand blob capacity for rollups while also raising the ceiling for mainnet execution. That approach also gives developers another way to lower pressure on rollups during periods when users prefer direct mainnet settlement.
Core developers are testing Glamsterdam on devnets, and the Sepolia fork now targets October 6, 2026. If the upgrade reaches mainnet as planned, Ethereum will enter its next scaling phase with more work happening both above Layer 1 and directly inside it.