Ethereum is preparing for the most consequential protocol upgrade since The Merge. Codenamed Glamsterdam, the hard fork scheduled for mid-2026 will introduce parallel transaction execution, enshrine proposer-builder separation into the consensus layer, and raise the gas limit from 60 million to a...
"If you create a 10,000 TPS EVM where its connection to L1 is mediated by a multisig bridge, then you are not scaling Ethereum." — Vitalik Buterin, Ethereum Co-Founder, February 2026
Ethereum is preparing for the most consequential protocol upgrade since The Merge. Codenamed Glamsterdam, the hard fork scheduled for mid-2026 will introduce parallel transaction execution, enshrine proposer-builder separation into the consensus layer, and raise the gas limit from 60 million to as high as 200 million — a 3.3x expansion that would unlock an order-of-magnitude increase in Layer 1 throughput.
The timing is not accidental. On February 3, 2026, Vitalik Buterin declared that the original rollup-centric roadmap "no longer makes sense," citing slower-than-expected Layer 2 decentralization and Ethereum's newfound ability to scale directly on L1. Ten days later, Ethereum Foundation co-executive director Tomasz Stańczak — the architect behind the Glamsterdam gas limit roadmap — announced his departure, injecting leadership uncertainty into the most ambitious upgrade cycle in Ethereum's history.
This report examines the three pillars of Glamsterdam: EIP-7732 (Enshrined Proposer-Builder Separation), EIP-7928 (Block-Level Access Lists for parallel execution), and EIP-7904 (gas cost repricing) — and analyzes what a successful deployment means for Ethereum's competitive position, its $45 billion Layer 2 ecosystem, and the $240 billion in value secured by the network.
For four years, "rollup-centric" was gospel. Ethereum's official scaling roadmap, codified in late 2021, positioned the base layer as a secure settlement and data availability substrate while Layer 2 rollups handled user-facing throughput. The logic was sound: keep L1 simple, let execution proliferate above it.
But the execution of that vision exposed structural problems. By early 2026, the dominant L2s — Arbitrum, Optimism, Base, zkSync — still rely on multisig-controlled upgrade mechanisms. Stage 2 decentralization, where rollups operate with fully trustless fraud or validity proofs, has proven "slower and more difficult" than expected, as Buterin acknowledged in his February blog post.
Meanwhile, Ethereum L1 itself became unexpectedly cheap. Post-Dencun blob space and the Fusaka upgrade (December 2025) kept base layer fees near zero for most users. The question shifted: if L1 can handle more load and L2s haven't decentralized their security models, why not scale L1 itself?
Glamsterdam is Buterin's answer. The upgrade redefines what "scaling Ethereum" means: creating "large quantities of block space that is backed by the full faith and credit of Ethereum," where activity is "guaranteed to be valid, uncensored, unreverted, untouched, as long as Ethereum itself functions." Anything less, he argues, is not Ethereum scale — it's a sidechain with extra steps.
Today, 90% of Ethereum blocks are constructed by just two entities: Beaverbuild and Titan. This duopoly emerged because the current proposer-builder separation (PBS) infrastructure — MEV-Boost relays operated by Flashbots and others — exists entirely off-chain. Builders compete for block construction, but the relay infrastructure that mediates the auction is centralized, trusted, and opaque.
The consequences are severe. Exclusive orderflow agreements between builders and trading venues have tripled builder concentration since 2022. A handful of relays can censor transactions. And the entire MEV supply chain — worth billions of dollars annually — depends on software that sits outside the protocol's security guarantees.
EIP-7732 moves this entire mechanism on-chain. Under enshrined PBS (ePBS), the consensus layer itself manages the commit-reveal flow between proposers and builders:
This is not an optimization. It is an architectural change that removes the dependency on trusted intermediaries from Ethereum's block production pipeline. If Flashbots disappeared tomorrow, block production under ePBS would continue uninterrupted.
The implications extend to censorship resistance. Today, a relay can refuse to propagate blocks containing sanctioned addresses. Under ePBS, censorship must occur at the builder level, and because builders compete in an on-chain auction, the economic cost of systematic censorship increases substantially.
Ethereum's EVM currently processes transactions sequentially. Every transaction in a block executes one after another, regardless of whether they touch the same state. This creates a fundamental throughput bottleneck: no matter how much gas a block contains, execution speed is limited by the single-threaded pipeline.
EIP-7928 introduces Block-Level Access Lists (BALs), a mechanism that records all accounts and storage slots accessed during block execution. By knowing in advance which state each transaction will touch, nodes can identify independent transactions and execute them simultaneously.
The technical elegance is in the enforcement. Unlike EIP-2930, which introduced optional transaction-level access lists that nobody uses (because they're not enforced), BALs are mandatory at the block level. Builders must include them. This gives every node in the network a map of the block's state dependencies, enabling:
The overhead is modest. Analysis from Ethereum Research estimates BALs add approximately 70 KiB to average block size — a negligible increase relative to the execution speedup gained. Multiple client teams, including Nethermind and Besu, are actively prototyping implementations.
The critical nuance: BALs do not require changes to the consensus layer. This is a pure execution-layer upgrade, which reduces coordination complexity and lowers the risk of cross-layer bugs during deployment.
Ethereum's gas pricing model has accumulated years of misalignment between what operations actually cost and what the protocol charges for them. The Gas Cost Estimator project — a benchmarking initiative spanning seven major EVM implementations — found high variance in execution time per unit of gas across opcodes. Some operations are dramatically overpriced; others are underpriced relative to their actual computational burden.
EIP-7904 corrects this using empirical data. The proposal adjusts gas costs for opcodes based on measured client benchmarks, adopting a conservative approach that prioritizes decreasing costs for demonstrably overpriced operations while avoiding reductions that could compromise security.
Key changes include a simplified memory expansion cost formula that reduces gas overhead for the first 22 words of memory (covering most contract interactions) and recalibrated costs for common computational operations. The net effect: more useful computation per unit of gas, which compounds with the gas limit increase to dramatically expand effective block capacity.
This isn't glamorous work, but it's economically essential. When you triple the gas limit, mispriced opcodes become exploitable attack vectors. Getting the pricing right is a precondition for the scaling Glamsterdam promises.
The gas limit increase is the most visible — and most aggressive — component of Glamsterdam. Before his departure, Stańczak outlined a phased approach at the Bankless Summit: 100 million gas in the first half of 2026, doubling to 200 million after ePBS proves stable, with 300 million possible before year-end.
To understand the magnitude: Ethereum's gas limit has increased gradually from around 30 million to 60 million over several years. Glamsterdam proposes to triple it in a single upgrade cycle.
The enabler is parallel execution. Without BALs, a 200-million-gas block would simply take three times longer to execute sequentially, potentially breaking block time targets and threatening network stability. With parallel execution, the increased gas is distributed across concurrent processing lanes, maintaining acceptable execution times.
The target is 10,000 transactions per second on Layer 1 — a number that would make Ethereum competitive with purpose-built high-throughput chains like Solana and Sui, but with the full security guarantees of Ethereum's validator set.
For the DeFi ecosystem, the implications are immediate. Higher gas limits support more complex smart contract interactions per block, reduce competition for block space during volatility spikes, and could permanently lower gas fees for users even as on-chain activity grows.
Glamsterdam creates an existential question for Layer 2s: if L1 can handle 10,000 TPS at low cost, what is the L2 value proposition?
The answer varies by rollup maturity. For Base, Coinbase's L2, the upgrade is actually welcome. Base has identified six promising non-headliner EIPs in the Glamsterdam proposal set, particularly around blob scaling that increases the data availability capacity L2s depend on. More blob space means lower settlement costs for rollups, even as L1 becomes more capable.
But Buterin's February remarks redefine the L2 competitive landscape. He explicitly stated that L2s whose "connection to L1 is mediated by a multisig bridge" are not scaling Ethereum. This disqualifies most current rollup deployments from claiming Ethereum's security guarantees.
Going forward, Buterin suggested L2s should provide distinct value: privacy features, application-specific design, ultra-fast sub-second confirmations, or non-financial use cases. Generic EVM chains that merely offer cheaper versions of L1 face a shrinking moat as Glamsterdam delivers cheap execution on L1 itself.
The economic realignment is significant. L2 tokens derive value partly from the narrative that they are essential scaling infrastructure. If L1 scaling reduces that essentiality, token valuations must be re-anchored to the distinct utility each L2 actually provides.
On February 13, 2026, Stańczak announced his departure from the Ethereum Foundation co-executive director role, effective end of February. Bastian Aue will take an interim leadership position.
The timing raises questions. Stańczak was the operational force behind the aggressive 2026 scaling roadmap. His push for a 10x L1 throughput increase within a single calendar year was ambitious even by Ethereum's standards. With the primary champion stepping down months before Glamsterdam's target deployment, execution risk increases.
However, the technical work is already well advanced. EIP-7732 and EIP-7928 have both been selected as Glamsterdam headliners through the formal All Core Devs process. Multiple client teams have begun prototyping. The June 2026 target remains aspirational, but the EIP specifications are stable and the development pipeline is in motion regardless of leadership changes at the Foundation level.
The deeper concern is governance continuity. The Ethereum Foundation has faced criticism over organizational opacity and slow decision-making. A leadership transition during the most complex upgrade since The Merge tests whether the Foundation's technical processes are robust enough to function independent of individual leaders.
Glamsterdam represents a philosophical inflection point for Ethereum. For years, the network accepted throughput limitations on L1 as a necessary trade-off for decentralization, delegating execution to an increasingly complex L2 ecosystem. That trade-off is being renegotiated.
The combination of parallel execution, enshrined PBS, and a tripled gas limit is an assertion that Ethereum can scale without sacrificing the properties that make it valuable: trustless execution, censorship resistance, and credible neutrality. If Glamsterdam succeeds, Ethereum moves from a settlement layer that processes a few dozen transactions per second to a full-stack execution environment capable of 10,000 TPS — while maintaining the decentralization guarantees that alt-L1s have struggled to match.
The stakes are proportional. A failed or delayed deployment would vindicate critics who argue Ethereum's governance is too slow for a market that moves fast. A successful one would reshape the competitive landscape across the entire smart contract platform sector and force a fundamental repricing of L2 tokens, alt-L1 valuations, and the infrastructure bets that have defined crypto investing since 2021.
The code is being written. The gas limit targets are set. The question is whether Ethereum's institutions — Foundation, client teams, and community — can execute at the pace the market demands.