The Glamsterdam Ethereum Upgrade: Ethereum’s Most Ambitious Redesign Since the Merge

Ethereum isn’t just iterating—it’s rewriting its core operating assumptions. The Glamsterdam upgrade, targeted for the first half of 2026, signals a deliberate shift away from off-chain dependencies toward protocol-native coordination. That’s not incremental progress. It’s structural reform at the base layer, tackling three of Ethereum’s most persistent criticisms—high fees, slow throughput, and centralized block production—in a single hard fork.

If the Merge made Ethereum sustainable, Glamsterdam may make it usable at global scale.

Why Glamsterdam Matters Now

To understand what Glamsterdam is doing, it helps to understand what it’s fixing.

Since the Merge, Ethereum has made real progress on energy consumption, Layer 2 scaling, and data availability. But the base layer itself has been sitting on a set of structural problems that previous upgrades either deferred or addressed only partially. The most glaring among them:

MEV centralization. Roughly 80–90% of Ethereum blocks are currently routed through a handful of external relay systems under the MEV-Boost model. This means a tiny slice of off-chain middleware controls most of the network’s block production—an arrangement that creates censorship risks, cartelization pressure, and a fundamental dependency on trust in third parties.

Sequential execution. Ethereum processes transactions one by one in strict order. Even when two transactions have nothing to do with each other—touching entirely different contracts and accounts—they still queue up and wait their turn. Modern hardware has multiple cores sitting idle while the EVM grinds through each operation in sequence.

Gas fees. Complex DeFi interactions remain expensive, particularly for users of the base layer. The combination of low throughput, high demand, and inefficient gas pricing structures means users pay a premium just to participate.

Glamsterdam addresses all three in a coordinated dual-layer hard fork. Its name is a portmanteau of two layer-specific codenames: Amsterdam for the execution layer and Gloas for the consensus layer. Together they form the most aggressive structural overhaul Ethereum has seen since it stopped mining.

The Architectural Shift: From Off-Chain to Protocol-Native

The deeper story of Glamsterdam is about trust—specifically, about where Ethereum’s trust assumptions have been quietly living outside the protocol.

Under the current MEV-Boost model, validators outsource block construction to builders who deliver blocks through relay services that sit entirely outside Ethereum’s consensus rules. Validators trust relays not to tamper with block contents, but that trust has no cryptographic enforcement behind it. With three or four relays handling the majority of blocks, the attack surface is concentrated in a way that would have seemed alarming if it had been designed that way from the start.

Glamsterdam’s central argument is that credible neutrality cannot be outsourced. Those guarantees need to live in the protocol itself—not in the reputations of relay operators.

ePBS (EIP-7732): Ethereum’s Answer to MEV Centralization

The MEV-Boost relay model functions as a trusted intermediary between block builders and validators. Builders construct blocks, seal them, and submit bids to relays. Validators select the highest bid without seeing block contents—a process that works only because validators trust relays to deliver what was promised. Relays can fail, causing missed blocks. They can filter transactions, making censorship possible at scale with no protocol-level accountability. And because relay operation is technically demanding and economically marginal, the market has consolidated toward a handful of dominant operators.

EIP-7732 moves this entire market into Ethereum’s consensus layer. Builders submit cryptographically sealed bids committed in-protocol. Validators select the highest bid as before—but now the commitment is enforced by the same consensus rules that govern everything else. No external relay required. Builders become first-class protocol participants with registered identities, signed bids, and protocol-enforced accountability.

One remaining problem is delivery: once a builder wins an auction, what stops them from withholding block contents to exploit the gap between commitment and reveal? Glamsterdam addresses this through the Payload Timeliness Committee (PTC), a group of validators that enforces delivery deadlines. Strict payload schedules and slashing conditions close what researchers call the “free option” problem—the asymmetric advantage builders previously had by committing to a bid and then deciding whether to follow through.

ePBS doesn’t eliminate MEV. The underlying incentive—that some transaction orderings are more profitable than others—remains. What it does is commoditize MEV extraction, turning a privileged insider market into a transparent, protocol-enforced auction. MEV extraction is projected to fall by up to 70% as a result.

Parallel Execution and Block-Level Access Lists: Ethereum’s Throughput Breakthrough

Ethereum currently discovers which accounts and storage slots a transaction will touch only during execution—no advance information. Every transaction queues up sequentially because there’s no way to know whether two transactions conflict without actually running them. Multi-core hardware buys almost nothing, because the EVM is effectively single-threaded.

EIP-7928 fixes this by introducing Block-Level Access Lists (BALs)—structured declarations specifying exactly which storage slots and accounts each transaction will read or write, committed in the block header before execution begins. With that state-access map available upfront, Ethereum clients can identify which transactions are genuinely independent and run them simultaneously across multiple CPU cores. Transactions that share state dependencies get grouped into sequential lanes; everything else runs in parallel.

The practical result is that high-traffic DeFi contracts stop being network-wide bottlenecks—their congestion gets contained within their own execution lane. Devnet results show sync speeds roughly 5x faster than the current baseline, with a near-term throughput target of 1,000+ TPS and a path to 10,000 as the gas limit scales.

Gas Economics Rewritten: Why Fees Could Drop by Around 78%

Glamsterdam bundles several gas repricing EIPs alongside the headline changes. EIP-7904 introduces benchmarked gas pricing, tying operation costs more directly to actual resource consumption on reference hardware for a more predictable cost structure. EIP-8007 adjusts pricing to penalize state bloat while rewarding computationally efficient patterns—an economic nudge toward long-term sustainability rather than raw expansion.

The gas limit itself is set to increase in stages, targeting 100 million per block initially and 200 million once ePBS is fully operational, up from roughly 30–60 million today. Combined with parallel execution and repricing, devnet results show approximately 78.6% fee reductions for complex DeFi interactions. For a user currently paying $4–8 per swap on the base layer, that’s a meaningful shift—not a footnote.

Security, Censorship Resistance, and Inclusion Guarantees

As relay infrastructure consolidated, censorship at the block level became a genuine vulnerability. Relay operators responding to regulatory pressure could filter transactions from specific addresses before they reached validators—invisibly, with no protocol-level accountability.

Under ePBS, builders who exclude transactions become economically accountable. The protocol can observe builder behavior directly and apply penalties for unjustified non-inclusion. Glamsterdam also lays groundwork for forward inclusion lists, a mechanism allowing validators to assert that specific transactions must appear in the next block. Full realization of this comes with the Hegota upgrade later in 2026 through FOCIL (Fork-Choice Enforced Inclusion Lists). Glamsterdam’s version is a foundation—but it transforms censorship from a silent risk into a behavior the protocol can identify and punish.

Quantitative Impact: Before and After Glamsterdam

Risks, Tradeoffs, and Open Questions

Glamsterdam’s ambition comes with real risks. Moving block building and execution logic into the consensus layer creates a broader attack surface—more moving parts, more edge cases, and neither ePBS nor BALs have been stress-tested at mainnet scale. On-chain block auctions introduce game-theoretic attack vectors that relay systems didn’t have; the Payload Timeliness Committee addresses the most obvious ones, but the full range of adversarial strategies in a live market is hard to anticipate in devnet conditions.

Parallel execution also shifts the performance bottleneck from compute toward disk I/O, which could strain nodes running on modest hardware. And BAL adoption requires wallets and dApp teams to update their tooling before the parallel-execution benefits fully materialize.

Then there’s the timeline. The June 2026 target is aspirational. Over 25 additional EIPs are under consideration for inclusion, and the Base engineering team has flagged that adding FOCIL alongside ePBS could push the upgrade into Q3 or Q4. Ethereum is trading simplicity for scalability. Whether that trade holds depends on how well client teams implement these changes under real-world conditions.

The Bigger Picture: Glamsterdam as a Bridge to What Comes Next

Glamsterdam is explicitly positioned as infrastructure for the upgrades that follow it. Hegota, expected in the second half of 2026, will introduce FOCIL for full inclusion list enforcement and encrypted mempools. Both features are significantly easier to build in a world where the protocol can identify builders and hold them accountable—something ePBS makes possible for the first time.

BALs also serve as a prerequisite for stateless clients: nodes that verify blocks without storing the full Ethereum state. This would lower hardware requirements for running a node substantially, with direct implications for decentralization. Vitalik Buterin’s scaling framework is explicit on the sequencing: ePBS and BALs deliver 10–30x execution improvements now; ZK-EVMs deliver 1,000x later. Glamsterdam is what makes the longer-range path actually feasible.

Final Verdict: Is Glamsterdam Ethereum’s Inflection Point?

Glamsterdam is the first Ethereum upgrade to simultaneously reduce fees, increase throughput, and improve decentralization at the base layer. Previous hard forks targeted one dimension at a time. This one targets all three, with a projected 70% reduction in MEV extraction, roughly 78% lower fees for complex DeFi interactions, and a throughput leap from 15–30 TPS toward 1,000+.

The more important shift, though, is structural. By moving block building and execution coordination into the protocol itself, Glamsterdam eliminates the informal middleware that Ethereum’s credible neutrality had quietly been depending on. Whether it ships on schedule is an open question—the June 2026 target is aspirational and the scope is large. But as a statement of architectural intent, it’s the clearest signal in years that Ethereum’s base layer is still being actively built, not just maintained.