Ethereum Glamsterdam Upgrade: Gas Limit Surges to 200 Million, How to Achieve 10,000 TPS?

In early May 2026, over 100 Ethereum core contributors gathered in Longyearbyen within the Arctic Circle to complete a week-long Soldøgn interoperability conference under continuous daylight, and announced that the three major technical goals of the Glamsterdam upgrade are essentially in place: Gas cap limits locked at 200 million, ePBS stable in external Builder processes, and final confirmation of EIP-8037 re-pricing parameters. This upgrade is expected to go live on the mainnet around June 2026 and is widely regarded as the most significant performance enhancement for Ethereum since The Merge in 2022.

What does Ethereum’s Gas limit increase from 60 million to 200 million mean?

Ethereum’s current Gas limit is approximately 60 million, which has been gradually increased from 30 million through two upgrades, Pectra and Fusaka, since 2021. Glamsterdam raises it all at once to 200 million, meaning the amount of computational operations that can fit in a single block will increase by 3.3 times, and the network’s theoretical throughput will jump from about 1,000 TPS to roughly 10,000 TPS.

Simply increasing the Gas limit does not automatically mean the network can truly handle that load. If execution client performance cannot keep up, a higher Gas limit is just a theoretical number, and congestion may still occur during peak times. However, with the support of ePBS, EIP-8037, and Block-Level Access Lists, this Gas limit increase is backed by multiple layers of safeguards—from low-level execution to state storage.

How does ePBS achieve protocol-level restructuring of block production?

Enshrined Proposer-Builder Separation (ePBS) is the core architectural change in Glamsterdam. Its essence is the complete separation of block proposers and builders at the protocol level. Previously, this separation relied on external relays and third-party Builder networks, but ePBS embeds it into the consensus layer, removing the trust dependency on third parties.

ePBS introduces explicit deadlines for block building, payload reveal, and attestation, providing more room on the timeline for the execution layer. This means validators no longer need to handle complex block construction tasks simultaneously and can focus on validation; professional builders with high-performance computing capabilities can independently optimize block strategies. The protocol-embedded design also introduces Payload Timeliness Committees and dual deadline logic, increasing throughput while reducing bottlenecks during block validation.

How does Block-Level Access Lists enable parallel execution and performance uplift?

BAL’s effect is more akin to low-level optimization: by allowing clients to fetch the read-write set of a block in advance before execution, it enables transaction parallelism and batch I/O. This improvement does not directly increase maximum throughput but prioritizes enhancing the performance of the slowest execution path—after the Gas limit is significantly increased, node synchronization speed and state root calculation efficiency become more critical, and BAL is designed to address this.

Ethereum’s path toward parallelization can be understood as phased evolution: early focus on serial execution performance and storage structure optimization, mid-term testing of benchmarks at the development network level via BAL, and later transitioning toward more extensive high-parallelism transaction models. Glamsterdam does not turn Ethereum into a fully parallel blockchain overnight, but it lays the infrastructure foundation for more efficient parallel execution models.

How does EIP-8037 address state bloat and resource pricing imbalance?

The cost of significantly increasing the Gas limit is that state data will expand at a faster rate. Ethereum’s global state records all account balances and contract data; if uncontrolled, the state will become the largest burden for full node operation.

EIP-8037 replaces the previous dynamic per-state-byte pricing with a fixed cost_per_state_byte, increasing the Gas cost for creating new states. This prevents attackers and inefficient contracts from deploying at low costs that could explode state storage. This mechanism ensures that even if block capacity expands to 200 million, the marginal cost of adding new state data remains aligned with actual hardware storage costs, avoiding a scenario where “more work in blocks” directly leads to unsustainable database growth.

How does L1 execution layer expansion change the value and competitive landscape of L2?

In recent years, Ethereum’s expansion narrative has been centered on Rollup-centric scaling—moving most execution tasks to L2 networks, with the L1 mainnet focusing on providing high-security settlement. Glamsterdam signals a clear message: the boundary of mainnet execution capacity is being redefined.

L2 networks handle 95% to 99% of transactions in the Ethereum ecosystem, but L1 transfer fees have fallen to very low levels. After the Glamsterdam upgrade, L1 data settlement costs will further decrease, with Rollup fees expected to drop by about 70%. This benefits large L2 solutions in terms of cost, but also broadens the scenarios covered by the mainnet—many simple transactions that previously had to be done via L2 can now be more conveniently completed directly on L1.

For L2 projects, this presents both short-term gains and medium-term pressures. The cost reduction is immediate, but they must demonstrate their unique value and higher efficiency to the market; otherwise, the question of “why use L2 instead of executing directly on L1” will become a real user consideration.

How do the timeline and execution certainty of the Glamsterdam upgrade manifest?

Before the Soldøgn interoperability conference, many technical parameters of Glamsterdam were still under discussion. After a week of around-the-clock testing, the finalized specifications were validated on the glamsterdam-devnet-2 testnet, with end-to-end testing of external Builder pathways across clients, and multi-client development networks remained stable.

EIP-8061 was included in the upgrade, EIP-8080 was explicitly rejected, and EIP-8045 was scoped down to a limited window of proposer responsibilities. These decisions indicate the team has transitioned from “feasibility discussions” to “consolidated executable specifications.” Final parameters will be formally confirmed at the upcoming AllCoreDevs meeting, with mainnet activation expected around June 2026.

Will there be further expansion after this? What is the roadmap after Glamsterdam?

According to the Ethereum Foundation’s 2026 protocol priority updates, protocol work has been reorganized into three long-term tracks: Scale, Improve UX, and Harden the L1. Glamsterdam is a key milestone within the Scale track, not the endpoint. The industry generally believes that the 200 million Gas limit is not the ceiling for this expansion—after Glamsterdam, the Gas limit will continue to push higher.

Following Glamsterdam is the H