Ethereum Glamsterdam Upgrade: ePBS, 78% Gas Cuts and 10K TPS Explained

If paying $30 to move $50 worth of ETH ever made you question the point of crypto — you are not alone. Ethereum has been wrestling with that gap between promise and cost for years. Glamsterdam is the protocol's most ambitious answer yet: a coordinated hard fork targeting roughly 10,000 transactions per second and gas fees up to 78% lower. Below is what it actually changes, what context makes it make sense, and what risks are worth keeping in your peripheral vision while you wait.

What Is Glamsterdam?

Glamsterdam is Ethereum's next coordinated hard fork and the direct successor to the Fusaka upgrade. The name follows Ethereum's tradition of combining a Devcon host city with a star name: Gloas (consensus layer) + Amsterdam (execution layer). Core developers have described it as the most significant protocol change since The Merge — a characterization that reflects both the breadth of what it touches and the size of the performance jump it targets.

The stated goals are blunt: push effective base-layer throughput from roughly 1,000 TPS toward ~10,000 TPS, cut gas costs sharply, expand capacity for Layer 2 data, and do all of that without weakening the network's decentralization properties.

The Core Changes

EIP-7732: Enshrined Proposer-Builder Separation (ePBS)

Think of today's Ethereum block production as a relay race where an outside coordinator handles the handoff between the runner who decides which block gets proposed and the builder who actually assembles it. That outside coordinator today is an off-protocol relay, most commonly MEV-Boost. The arrangement works, but it means the network depends on infrastructure that sits outside the protocol itself.

ePBS removes that third party and bakes the handoff directly into Ethereum. The practical result: the block data-propagation window expands from roughly 2 seconds to about 9 seconds. That extra window is what unlocks higher throughput and gives the network room to carry more data blobs for rollups. It also reduces reliance on off-protocol infrastructure — a persistent concern among researchers tracking censorship resistance and validator fairness.

EIP-7928: Block-Level Access Lists (BALs)

Block-Level Access Lists work like a picking list published before a warehouse run. Right now, transactions execute largely in sequence: each step waits for the previous one to finish before starting. With BALs, a block declares upfront exactly which accounts and storage slots it will read or write. That declaration lets clients parallelize transaction execution and validation — processing multiple operations simultaneously rather than in a single-file queue.

Parallelization is one of the key levers for reaching higher TPS without requiring every network node to upgrade to significantly more powerful hardware.

Gas Repricing and Lower Fees

Glamsterdam realigns gas costs with the actual computational work each operation demands. Some historically inexpensive-to-compute operations have carried inflated gas costs for legacy reasons; others have been underpriced relative to the load they put on nodes. The repricing targets a correction of roughly 78.6% lower gas costs for both simple ETH transfers and complex smart-contract interactions.

The upgrade also targets a dramatic increase in the gas limit ceiling — discussions point toward a 200-million gas floor, up from roughly 60 million today. A higher gas limit means more computation fits in each block, which compounds the throughput gains from BALs and ePBS.

If you have ever checked Ethereum gas fees and decided a transaction was not worth sending, this repricing is aimed directly at that friction.

Why Glamsterdam Matters

The three changes above compound each other in ways that extend well beyond the base layer. ePBS's wider propagation window creates headroom for more data blobs — the mechanism rollups use to post transaction data back to Ethereum. More blob capacity means lower data-posting costs for Layer 2 networks, which flows through to cheaper transactions on rollups. This reinforces Ethereum's rollup-centric scaling roadmap and keeps L2 economics improving even before the next dedicated blob-focused fork.

For validators and ETH stakers, enshrining PBS at the protocol level reduces dependence on off-protocol relay networks. That matters for censorship resistance and for making block production more accessible to smaller validators who prefer not to rely on a third-party relay.

For dApp developers, the combination of lower gas costs and higher throughput makes the base layer more practical for use cases that currently cannot justify L1 costs. Applications pushed to Layer 2 purely because of L1 expense could find that calculus shifting after Glamsterdam activates.

Context: How We Got Here

Understanding where Glamsterdam sits requires a brief look at the sequence it belongs to — not to predict what comes next, but to see the pattern clearly.

Ethereum launched on Proof-of-Work: secure but energy-intensive and capacity-limited. The Beacon Chain (December 2020) and The Merge (September 2022) moved the network to Proof-of-Stake, cutting energy use dramatically without immediately expanding transaction throughput. EIP-1559 (August 2021) made gas pricing more predictable. The Dencun upgrade (March 2024) introduced proto-danksharding via EIP-4844 — adding blobs and reducing L2 data-posting costs significantly, the first hard fork specifically targeting rollup economics. Fusaka, the upgrade immediately before Glamsterdam, extended blob capacity further.

Glamsterdam is the next step in that chain: adding the ePBS and parallelization changes that earlier hard forks deliberately deferred to lay groundwork first. Shipping foundational changes before building on them has been a consistent pattern across Ethereum's roadmap.

It is also worth noting — as an observable fact, not a prediction — that Ethereum hard fork timelines have historically run behind initial targets. Shapella, Dencun, and several earlier upgrades each saw schedule shifts before landing on mainnet. The community has consistently preferred delay over shipping under-tested changes to a network carrying substantial value. That pattern is relevant context for reading any current timeline estimate.

Risks and Caveats

No upgrade of this scope ships without real risk. A few worth tracking:

• Timeline risk: Glamsterdam originally targeted June 2026. As of mid-June 2026, the window had moved to Q3–Q4 2026 (September–December), and EIP-7773 — which tracks the final scope — remained in Draft status with roughly ten EIPs listed for inclusion. Further slippage is possible; treat specific dates as tentative until the Ethereum Foundation confirms a mainnet activation epoch after successful testnet runs.
• Scope uncertainty: ePBS and BALs are confirmed in scope. Other proposals — such as 6-second slots — have been discussed but are not part of the current confirmed set. The final EIP bundle may look different from what is currently expected.
• Technical transition risk: Hard forks require all network clients to coordinate upgrades simultaneously. A client bug or coordination failure during the transition could cause a temporary chain split. Ethereum's multi-client architecture and testnet process are designed to catch these issues before mainnet, but the risk is real and worth acknowledging.
• L2 benefit timing: The full reduction in Layer 2 costs is unlikely to arrive in one step. Future BPO (Blob Parameter Only) forks are expected to incrementally raise blob capacity over time, meaning rollup economics improve gradually rather than all at once on Glamsterdam's activation date.
• No price guarantee: Protocol upgrades do not translate automatically into market price movements. Markets respond to many factors simultaneously, and volatility in crypto markets is high. Any market exposure should only involve amounts you can afford to lose. Always do your own research. (NFA)

Frequently Asked Questions

Does Glamsterdam affect my ETH wallet or holdings? No change to how you hold or send ETH. If anything, you would notice lower fees when you transact after the upgrade activates — no new wallet step required.

Will gas fees actually drop 78% after Glamsterdam? The ~78% figure is the targeted reduction from the gas repricing proposal, not a guarantee. Actual transaction fees also depend on network congestion and demand at the time. The repricing changes the baseline cost structure, not the variable demand component.

Is ePBS the same as MEV-Boost? They address a related problem at different layers. MEV-Boost is an off-protocol relay used today to separate proposers and builders. ePBS achieves the same separation inside the protocol itself, removing the need for a trusted external relay in that role and reducing the network's dependency on off-protocol infrastructure.

What happens if the timeline slips again? The upgrade does not activate until the Ethereum Foundation confirms a mainnet activation epoch after successful testnet runs. A further delay means the current network continues operating exactly as-is — no disruption to existing users, just a longer wait for the fee and throughput improvements.

How does Glamsterdam compare to The Merge? The Merge switched Ethereum's consensus mechanism from Proof-of-Work to Proof-of-Stake — a fundamental change to how the chain is secured. Glamsterdam targets throughput and fee structure rather than the consensus mechanism. The "biggest since The Merge" framing reflects the breadth and ambition of protocol changes, not that the two upgrades are identical in nature.

Bottom Line

Glamsterdam bundles three interlocking changes — ePBS, Block-Level Access Lists, and gas repricing — into Ethereum's most ambitious hard fork in years. The stated targets are significant: a path toward ~10,000 TPS and roughly 78% lower fees. The caveats are equally real: the timeline has already moved once, the final EIP list is not yet locked, and hard forks carry technical transition risk that cannot be entirely engineered away.

What is observable is that Ethereum has executed consistently against its rollup-centric scaling roadmap across multiple upgrade cycles. Glamsterdam is the next step in that sequence — more throughput, cheaper fees, reduced off-protocol dependencies. Whether it ships on the current schedule or after another adjustment, the direction it represents has been consistent since The Merge.