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

Ethereum is not simply iterating—it is rewriting its core working assumptions. The Glamsterdam improve, focused for the primary half of 2026, alerts a deliberate shift away from off-chain dependencies towards protocol-native coordination. That is not incremental progress. It is structural reform on the base layer, tackling three of Ethereum’s most persistent criticisms—excessive charges, sluggish throughput, and centralized block manufacturing—in a single onerous fork.

If the Merge made Ethereum sustainable, Glamsterdam could make it usable at world scale.

Why Glamsterdam Issues Now

To grasp what Glamsterdam is doing, it helps to grasp what it is fixing.

For the reason that Merge, Ethereum has made actual progress on vitality consumption, Layer 2 scaling, and information availability. However the base layer itself has been sitting on a set of structural issues that earlier upgrades both deferred or addressed solely partially. Probably the most obvious amongst them:

MEV centralization. Roughly 80–90% of Ethereum blocks are at the moment routed by a handful of exterior relay methods underneath the MEV-Enhance mannequin. This implies a tiny slice of off-chain middleware controls many of the community’s block manufacturing—an association that creates censorship dangers, cartelization stress, and a basic dependency on belief in third events.

Sequential execution. Ethereum processes transactions one after the other in strict order. Even when two transactions don’t have anything to do with one another—touching solely completely different contracts and accounts—they nonetheless queue up and wait their flip. Trendy {hardware} has a number of cores sitting idle whereas the EVM grinds by every operation in sequence.

Gasoline charges. Complicated DeFi interactions stay costly, significantly for customers of the bottom layer. The mixture of low throughput, excessive demand, and inefficient gasoline pricing buildings means customers pay a premium simply to take part.

Glamsterdam addresses all three in a coordinated dual-layer onerous fork. Its identify is a portmanteau of two layer-specific codenames: Amsterdam for the execution layer and Gloas for the consensus layer. Collectively they type probably the most aggressive structural overhaul Ethereum has seen because it stopped mining.

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

The deeper story of Glamsterdam is about belief—particularly, about the place Ethereum’s belief assumptions have been quietly residing exterior the protocol.

Below the present MEV-Enhance mannequin, validators outsource block building to builders who ship blocks by relay providers that sit solely exterior Ethereum’s consensus guidelines. Validators belief relays to not tamper with block contents, however that belief has no cryptographic enforcement behind it. With three or 4 relays dealing with nearly all of blocks, the assault floor is concentrated in a manner that will have appeared alarming if it had been designed that manner from the beginning.

Glamsterdam’s central argument is that credible neutrality can’t be outsourced. These ensures have to reside within the protocol itself—not within the reputations of relay operators.

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

The MEV-Boost relay mannequin features as a trusted middleman between block builders and validators. Builders assemble blocks, seal them, and submit bids to relays. Validators choose the very best bid with out seeing block contents—a course of that works solely as a result of validators belief relays to ship what was promised. Relays can fail, inflicting missed blocks. They will filter transactions, making censorship attainable at scale with no protocol-level accountability. And since relay operation is technically demanding and economically marginal, the market has consolidated towards a handful of dominant operators.

EIP-7732 strikes this complete market into Ethereum’s consensus layer. Builders submit cryptographically sealed bids dedicated in-protocol. Validators choose the very best bid as earlier than—however now the dedication is enforced by the identical consensus guidelines that govern all the things else. No exterior relay required. Builders change into first-class protocol members with registered identities, signed bids, and protocol-enforced accountability.

One remaining drawback is supply: as soon as a builder wins an public sale, what stops them from withholding block contents to take advantage of the hole between dedication and reveal? Glamsterdam addresses this by the Payload Timeliness Committee (PTC), a gaggle of validators that enforces supply deadlines. Strict payload schedules and slashing situations shut what researchers name the “free possibility” drawback—the uneven benefit builders beforehand had by committing to a bid after which deciding whether or not to observe by.

ePBS would not remove MEV. The underlying incentive—that some transaction orderings are extra worthwhile than others—stays. What it does is commoditize MEV extraction, turning a privileged insider market right into a clear, protocol-enforced public sale. MEV extraction is projected to fall by as much as 70% because of this.

Parallel Execution and Block-Degree Entry Lists: Ethereum’s Throughput Breakthrough

Ethereum at the moment discovers which accounts and storage slots a transaction will contact solely throughout execution—no advance info. Each transaction queues up sequentially as a result of there is no method to know whether or not two transactions battle with out really working them. Multi-core {hardware} buys virtually nothing, as a result of the EVM is successfully single-threaded.

EIP-7928 fixes this by introducing Block-Degree Entry Lists (BALs)—structured declarations specifying precisely which storage slots and accounts every transaction will learn or write, dedicated within the block header earlier than execution begins. With that state-access map obtainable upfront, Ethereum purchasers can establish which transactions are genuinely impartial and run them concurrently throughout a number of CPU cores. Transactions that share state dependencies get grouped into sequential lanes; all the things else runs in parallel.

The sensible result’s that high-traffic DeFi contracts cease being network-wide bottlenecks—their congestion will get contained inside their very own execution lane. Devnet outcomes present sync speeds roughly 5x quicker than the present baseline, with a near-term throughput goal of 1,000+ TPS and a path to 10,000 because the gasoline restrict scales.

Gasoline Economics Rewritten: Why Charges May Drop by Round 78%

Glamsterdam bundles a number of gasoline repricing EIPs alongside the headline adjustments. EIP-7904 introduces benchmarked gasoline pricing, tying operation prices extra on to precise useful resource consumption on reference {hardware} for a extra predictable value construction. EIP-8007 adjusts pricing to penalize state bloat whereas rewarding computationally environment friendly patterns—an financial nudge towards long-term sustainability moderately than uncooked growth.

The gasoline restrict itself is about to extend in phases, concentrating on 100 million per block initially and 200 million as soon as ePBS is absolutely operational, up from roughly 30–60 million right now. Mixed with parallel execution and repricing, devnet outcomes present roughly 78.6% payment reductions for complicated DeFi interactions. For a consumer at the moment paying $4–8 per swap on the bottom layer, that is a significant shift—not a footnote.

Safety, Censorship Resistance, and Inclusion Ensures

As relay infrastructure consolidated, censorship on the block degree grew to become a real vulnerability. Relay operators responding to regulatory stress might filter transactions from particular addresses earlier than they reached validators—invisibly, with no protocol-level accountability.

Below ePBS, builders who exclude transactions change into economically accountable. The protocol can observe builder habits straight and apply penalties for unjustified non-inclusion. Glamsterdam additionally lays groundwork for ahead inclusion lists, a mechanism permitting validators to say that particular transactions should seem within the subsequent block. Full realization of this comes with the Hegota improve later in 2026 by FOCIL (Fork-Selection Enforced Inclusion Lists). Glamsterdam’s model is a basis—however it transforms censorship from a silent danger right into a habits the protocol can establish and punish.

Quantitative Affect: Earlier than and After Glamsterdam

Dangers, Tradeoffs, and Open Questions

Glamsterdam’s ambition comes with actual dangers. Shifting block constructing and execution logic into the consensus layer creates a broader assault floor—extra transferring components, extra edge circumstances, and neither ePBS nor BALs have been stress-tested at mainnet scale. On-chain block auctions introduce game-theoretic assault vectors that relay methods did not have; the Payload Timeliness Committee addresses the obvious ones, however the full vary of adversarial methods in a reside market is tough to anticipate in devnet situations.

Parallel execution additionally shifts the efficiency bottleneck from compute towards disk I/O, which might pressure nodes working on modest {hardware}. And BAL adoption requires wallets and dApp groups to replace their tooling earlier than the parallel-execution advantages absolutely materialize.

Then there’s the timeline. The June 2026 goal is aspirational. Over 25 further EIPs are into consideration for inclusion, and the Base engineering group has flagged that including FOCIL alongside ePBS might push the improve into Q3 or This autumn. Ethereum is buying and selling simplicity for scalability. Whether or not that commerce holds relies on how properly shopper groups implement these adjustments underneath real-world situations.

The Larger Image: Glamsterdam as a Bridge to What Comes Subsequent

Glamsterdam is explicitly positioned as infrastructure for the upgrades that observe it. Hegota, anticipated within the second half of 2026, will introduce FOCIL for full inclusion record enforcement and encrypted mempools. Each options are considerably simpler to construct in a world the place the protocol can establish builders and maintain them accountable—one thing ePBS makes attainable for the primary time.

BALs additionally function a prerequisite for stateless purchasers: nodes that confirm blocks with out storing the total Ethereum state. This could decrease {hardware} necessities for working a node considerably, with direct implications for decentralization. Vitalik Buterin’s scaling framework is specific on the sequencing: ePBS and BALs ship 10–30x execution enhancements now; ZK-EVMs ship 1,000x later. Glamsterdam is what makes the longer-range path really possible.

Ultimate Verdict: Is Glamsterdam Ethereum’s Inflection Level?

Glamsterdam is the primary Ethereum improve to concurrently scale back charges, enhance throughput, and enhance decentralization on the base layer. Earlier onerous forks focused one dimension at a time. This one targets all three, with a projected 70% discount in MEV extraction, roughly 78% decrease charges for complicated DeFi interactions, and a throughput leap from 15–30 TPS towards 1,000+.

The extra necessary shift, although, is structural. By transferring block constructing and execution coordination into the protocol itself, Glamsterdam eliminates the casual middleware that Ethereum’s credible neutrality had quietly been relying on. Whether or not it ships on schedule is an open query—the June 2026 goal is aspirational and the scope is giant. However as a press release of architectural intent, it is the clearest sign in years that Ethereum’s base layer remains to be being actively constructed, not simply maintained.