Ethereum (ETH) rehearsed its next overhaul on Oct. 6. Few people outside developer channels noticed. Yet the Sepolia testnet now runs rules that change who builds blocks, how gas is priced and how much work a block can carry.
This explainer unpacks Glamsterdam's 18 core changes, led by enshrined proposer-builder separation, block-level access lists and gas repricing, and what they mean for fees, MEV, rollups and a planned jump to a 200 million gas limit.
- Glamsterdam went live on Ethereum's Sepolia testnet on Oct. 6, 2026, and developers have set Oct. 26 for the Hoodi testnet. No mainnet date has been agreed, although a draft scheduling proposal lists early December options.
- The upgrade moves the proposer-builder handoff into the protocol, adds block-level access lists for parallel processing and reprices how Ethereum charges for creating and reading state.
- Developers aligned on a 200 million gas limit floor after Glamsterdam, more than three times today's 60 million default, though the effect on fees depends on demand and validator adoption.
What Is Ethereum's Glamsterdam Upgrade, and Why Does It Matter?
Glamsterdam is the network upgrade that follows Fusaka, and its name joins two halves. The execution layer side is called Amsterdam, after the 2022 Devconnect host city, while the consensus side is Gloas, named after a star.
The Ethereum Foundation published its testnet announcement on Sept. 28. It scheduled activation on Sepolia at epoch 353,024, slot 11,296,768, at 13:53:36 UTC on Oct. 6.
The fork landed on time.
Independent Ethereum researcher Christine D. Kim reported on Oct. 9 that the Sepolia upgrade "went smoothly." Developers then scheduled Glamsterdam for Hoodi, the second and final public testnet, on Monday, Oct. 26, with client releases expected by Oct. 16.
The scope sits in a tracking proposal, EIP-7773. It schedules 18 core changes, supported by five networking proposals and two informational ones.
Mainnet is a different story. The foundation said Hoodi and mainnet dates "have not yet been decided," and ethereum.org still lists only a fourth-quarter 2026 target.
A draft scheduling proposal on GitHub floats Dec. 1, Dec. 8 and Dec. 9 as potential mainnet slots, but developers have not agreed on any of them.
Why does it matter? For years, Ethereum pushed most scaling work onto rollups and kept its base layer conservative, so Glamsterdam marks a visible change in emphasis.
The foundation frames the upgrade as the next step in its plan to scale Layer 1. Its headline features exist mainly to let the base chain safely process far more activity, while pricing gas closer to the real cost of execution and state growth.
Also Read: ChatGPT’s Intelligent UI Puts Charts And Calculators In Answers For 1.2B Users
How Does Glamsterdam Differ From Pectra and Fusaka?
Ethereum's last two upgrades were mostly about staking and data. Glamsterdam is about execution, the part of the system that actually runs transactions.
Pectra went live on May 7, 2025, at epoch 364,032. It let ordinary wallets delegate to smart contract code through EIP-7702, raised the maximum validator balance from 32 to 2,048 Ether and lifted the blob target to six, with a maximum of nine.
Fusaka activated on Dec. 3, 2025, at 21:49:11 UTC. Its headline feature was PeerDAS, or EIP-7594, which lets nodes verify blob data by sampling small pieces rather than downloading everything.
Fusaka also introduced blob-parameter-only forks, small configuration upgrades that change blob capacity without bundling other features. Two followed, on Dec. 9, 2025, and Jan. 7, 2026.
The foundation's January Checkpoint update noted that Ethereum now targets 14 blobs per block and allows a maximum of 21. It called that a 2.3-fold increase in Layer 2 data space compared with the period before Fusaka.
On the execution side, Fusaka capped individual transactions at roughly 16.8 million gas and set the default block gas limit in client software at 60 million.
Glamsterdam changes the order of business. Rather than adding more room for rollup data, it rebuilds how blocks are produced, validated and priced, so that the 60 million figure can rise much further.
Pectra and Fusaka widened the pipes for Layer 2s, while Glamsterdam rewires the engine room of Layer 1.
Also Read: Can TRON Shield $95B In Stablecoins From Quantum Attacks? Sun Says It’s Ready
What Is Enshrined Proposer-Builder Separation, or ePBS?
Today, most Ethereum blocks are not assembled by the validators who propose them. Specialized builders compete to construct the most profitable block, and a validator running MEV-Boost software from Flashbots picks the highest bid.
The handoff relies on relays, which act as trusted middlemen. A relay holds the builder's block, shows the proposer only the header and the bid, and releases the full contents once the proposer signs.
That arrangement works, but it sits entirely outside the protocol. If a relay fails, misbehaves or censors, Ethereum's consensus rules offer no recourse.
EIP-7732 moves the exchange into Ethereum itself. It treats builders as protocol participants and gives validators a new duty, called payload timeliness attestations.
Under the new flow, a proposer includes a builder's signed commitment to an execution payload in the beacon block, and the builder reveals the payload afterward. The protocol handles the payment to the proposer, which reduces reliance on trusted middleware.
A subset of validators forms the payload timeliness committee. Its members attest to whether the builder revealed its payload on time and whether the matching blob data was available.
The change also splits consensus validation from execution validation, and that is where scaling enters the picture.
The ethereum.org roadmap page says the split widens the window for spreading data across the network from 2 seconds to about 9 seconds. More time means validators can safely handle much bigger payloads.
Relays will not vanish overnight. The same page notes that builders and proposers may keep using off-protocol relays or middleware for advanced features the core protocol does not yet support.
Also Read: Claude Mythos 5 Reached Government Websites, So Anthropic Cut The Cord On Tests
How Could Glamsterdam Change MEV and the Economics of Ethereum Validators?
Maximum extractable value, or MEV, is the profit available from ordering, including or excluding transactions. Through MEV-Boost, much of it flows to validators as builder bids, but the market behind those bids is highly concentrated.
Flashbots' Relayscan dashboard showed Titan building 52.6% of MEV-Boost blocks in the 24 hours to early Oct. 12. Quasar built 18.5% and BuilderNet 13.3%, leaving three builders with roughly 84% of the flow.
ePBS does not abolish MEV, and it does not automatically break up builder concentration. What it changes is the trust model, since a proposer's payment no longer depends on a relay honoring its promises.
For stakers, the benefit shows up at the margin. A solo validator gains a protocol-enforced way to sell block space, while builders must post a stake of their own to take part.
Stakers also get faster exits. Under EIP-8061, exit capacity scales with the total amount of staked Ether, which ethereum.org says roughly quadruples it at current levels and doubles the room for validator consolidations.
The trade-off is a shorter safety window for nodes that go offline. Ethereum.org says the weak subjectivity period, after which a returning node needs a recent trusted checkpoint, falls from about 15.7 days to about 7 days.
Developers at the Soldøgn interop in Svalbard, a spring gathering of just over 100 core contributors, flagged two ePBS questions that remain contentious. One is how to protect a design that asks builders for only 1 ETH of stake against Sybil attacks on the peer-to-peer network.
That worry resurfaced before Sepolia. On a Sept. 17 consensus call, Prysm developer Potuz warned that free test Ether would let an attacker win auctions with throwaway builders and then withhold payloads: "I can just spin up a thousand builders."
Early Sepolia data points to a thin market. A community summary of the Oct. 8 execution-layer call cited EthPandaOps as saying only four or five builders had been active there, with major builders absent and many blocks still built locally.
ePBS only proves its value when real builders compete for blocks, which makes Hoodi the more telling test.
Also Read: iPhone Duo May Say Hello To Buyers Personally When $1,999 Foldable First Boots
What Are Block-Level Access Lists, and How Do They Improve Transaction Processing?
Ethereum clients today largely process a block's transactions one after another. They cannot know in advance which accounts or storage slots a transaction will touch, so running two at once risks collisions.
EIP-7928 introduces block-level access lists, or BALs, which record every account and storage slot a block touches, along with the state changes each transaction makes. Builders produce the list, and the block header carries a hash of it.
With that map in hand, clients can do several jobs in parallel:
- Read the required state from disk at the same time, instead of discovering it step by step.
- Validate transactions that do not overlap on separate CPU cores.
- Compute the post-block state root more efficiently.
- Update state from the recorded results without replaying every transaction, which speeds up syncing.
A companion networking change, EIP-8159, lets nodes exchange these lists with each other. Another, EIP-8189, uses BAL data to heal state during snap sync.
At Soldøgn, client teams benchmarked each BAL optimization behind its own feature flag. The aim was to speed up each client's slowest paths first, because lifting the weakest implementation raises the safe gas limit for everyone.
BALs and ePBS work as a pair. ePBS gives execution more time inside each slot, and BALs let clients use that time far more efficiently.
Also Read: Bezos Says AI Could Mean Three-Day Workweeks After Amazon Shed 30,000 Jobs
How Will Glamsterdam Change Ethereum's Gas Accounting and Execution Costs?
Gas prices for state operations were last overhauled in the Berlin upgrade of 2021. Since then, Ethereum's state has grown significantly, and recent gas limit increases have sped up that growth.
A foundation post for smart contract developers explained on Aug. 24 that the new schedule supports "roughly a 3x increase in base throughput." Repricing these operations, it said, is a prerequisite for raising the gas limit further.
Two EIPs do the heavy lifting. EIP-8037 raises and separately meters the cost of creating state, while EIP-8038 raises the cost of reading and writing it through operations such as SLOAD, SSTORE and cold account access.
EIP-8037 sets a fixed price of 1,530 gas per new byte of state, derived from a target of 120 GiB of state growth a year at a reference gas limit of 150 million.
Under that rate, a new storage slot costs 97,920 units of state gas instead of 20,000 today, and a new account costs 183,600 instead of 25,000. Deploying a maximum-size 24 KiB contract becomes roughly eight times more expensive.
State gas also gets its own meter. The per-transaction cap introduced in Fusaka now applies only to execution gas, and a block counts whichever of the two dimensions is closer to its limit.
The rest of the gas package is narrower, but it adds up:
- EIP-7778 stops gas refunds from counting against the block gas limit, though users still receive refunds on what they pay.
- EIP-7976 raises the calldata floor to 64 gas per byte for data-heavy transactions.
- EIP-7981 charges access list data at the same 64 gas per byte.
- EIP-2780 rebuilds intrinsic transaction gas around actual resource use, with a 12,000 gas base cost.
- EIP-7954 raises the maximum contract size from 24 KiB to 64 KiB.
The current EIP-2780 draft keeps a plain Ether transfer to an existing account at 21,000 gas, while a transfer that creates a new account adds 183,600 units of state gas. Ethereum.org's Glamsterdam page, last updated Aug. 6, still describes transfers becoming up to 71% cheaper.
The numbers may still move. Several of these proposals, including EIP-2780 and EIP-8037, remain in Last Call review until Nov. 1, and ethereum.org notes that scope can still change before mainnet.
Also Read: Zcash Has A January Quantum Target For 70% Of ZEC, What It Lacks Is A Date
Could the Upgrade Increase Layer 1 Throughput and Reduce Transaction Fees?
Throughput is the payoff. At Soldøgn, developers aligned on a "credible 200M post-Glamsterdam gas limit floor," roughly 3.3 times today's 60 million default.
The number rests on three supports. ePBS gives execution more time, BAL optimizations give clients headroom, and EIP-8037 keeps a larger gas limit from producing runaway state growth.
Sepolia already runs a 200 million limit, scheduled at the fork epoch under EIP-8261, which turns the gas limit into a fork-scheduled parameter.
Client code for Hoodi schedules the same 200 million figure. Call summaries say mainnet will follow the same fork-epoch approach, but no mainnet figure or date has been formally announced.
Will fees fall? Ethereum.org says Glamsterdam "will most likely reduce fees for everyday users." More block space at the same demand should push prices down, but that outcome depends on how much new activity fills the room.
State-heavy actions, such as creating new accounts or storage slots, will cost more gas, so the upgrade redistributes costs as much as it cuts them.
Also Read: Kalshi Faces NFL At The Supreme Court As $1.8B Rides On One Football Sunday
What Must Wallets, Smart Contracts, Infrastructure Providers and Validators Prepare For?
For ordinary holders, the answer is nothing. The foundation says mainnet users do not need to act, and ethereum.org warns that anyone telling people to "upgrade" their ETH is running a scam.
Everyone else has homework, and the repricing is the biggest item. Replaying historical mainnet transactions revealed a small set of contracts that could break or degrade, typically because they rely on hardcoded gas assumptions.
The foundation said a simple gas-limit increase fixes most flagged issues, and that the large majority of smart contracts are unaffected.
The checklist drawn from foundation guidance and client teams looks like this:
- Smart contract developers should check addresses in the foundation's affected-contracts tool and test fixes on the Platåberget testnet, watching for Solidity's fixed 2,300 gas stipend, hardcoded gas values and logic that branches on remaining gas.
- Wallets, RPC providers and node tooling teams must update gas estimation, because cached gas constants will underestimate costs and cause failed transactions.
- Exchanges, bridges and indexers can use the new logs from EIP-7708, which make every non-zero Ether transfer and burn emit a standard event.
- Node operators must upgrade both execution and consensus clients before each activation, and validators must update the beacon node and the validator client.
- Staking operators and builder teams should review the new ePBS duties and interfaces, including the payload timeliness committee, before each fork.
Sepolia-ready execution clients included Geth 1.17.7, Nethermind 2.1.0, Besu 26.9.0, Erigon 3.7.1, Reth 2.7.0 and Ethrex 29.0.1. On the consensus side, the list included Lighthouse 8.3.0-rc.0, Prysm 7.2.1, Teku 26.9.1, Lodestar 1.49.0, Nimbus 26.10.0 and Grandine 3.0.0-rc.0.
Gas limit defaults show how much small details matter. Prysm's 7.2.0 release and Teku's Sepolia release both left validators at 60 million after the fork unless operators set 200 million themselves.
Prysm merged a fix on Oct. 2 so that validators without an explicit setting follow Sepolia's 200 million schedule, and the foundation now lists Prysm 7.2.1. Teku validators who want 200 million still have to set it with a configuration flag.
Also Read: AI Catastrophe Within 12 Months? OpenAI, Anthropic Reportedly Plan For Fallout
How Does Glamsterdam Affect the Relationship Between Ethereum Layer 1 and Layer 2 Networks?
Glamsterdam does not abandon rollups. The official roadmap frames the upgrade as advancing both its Scale L1 and Scale Blobs objectives.
The connection is mechanical. Because ePBS widens the propagation window, ethereum.org says Ethereum can safely carry much more data, including more blobs for rollups, without stressing the network.
Glamsterdam itself does not set new blob numbers. Since Fusaka, blob capacity changes through separate blob-parameter-only forks, so data capacity for rollups can keep growing on its own schedule.
Developers are in no rush on that front. The January Checkpoint said a third such fork was not a priority until blob usage catches up with the existing increases.
Two smaller features help Layer 2s directly. EIP-7997 places a deterministic factory contract at the same address on every participating EVM chain, so developers can deploy wallets and apps to identical addresses across networks.
The second, EIP-8070, introduces a sparse blobpool, a networking change to how nodes handle blob transactions.
The strategic tension is real, though. A cheaper, faster Layer 1 makes direct deployment on mainnet more attractive, which puts it in competition with the rollups that settle there.
Ethereum's answer, so far, is to scale both layers at once. The roadmap treats execution capacity and blob capacity as parallel goals, and Glamsterdam's design reflects that dual mandate rather than a choice between layers.
For rollup users, the practical effect is a sturdier, more trust-minimized settlement layer underneath them.
Also Read: XRP ETFs Draw $8M During Broad Crypto Fund Selloff Led By Bitcoin
What Comes After Glamsterdam in Ethereum's Scaling Roadmap?
The next fork is Hegotá. Like Glamsterdam, its name pairs an execution-layer city, Bogotá, with a consensus-layer star, Heze.
Its headliners are already locked. The foundation's Protocol cluster named EIP-7805, fork-choice enforced inclusion lists or FOCIL, for the consensus layer and EIP-8141, Frame Transactions, for the execution layer.
FOCIL was once a Glamsterdam candidate. Developers moved it out to limit the fork's scope, according to the foundation's January Checkpoint.
FOCIL lets multiple validators force builders to include valid transactions, which directly addresses the censorship risk that builder concentration creates. Frame transactions make validation and gas payment programmable, giving accounts a native path away from today's secp256k1 keys.
The cluster's companion tier list graded 62 proposed EIPs, drawing on 397 tier grades from a cluster of about 60 people. Two were rated must-ship, 15 expected to ship and 28 declined.
Timing is ambitious. The cluster said client teams could realistically begin Hegotá implementation in late Q4 2026, and its planning assumes Glamsterdam ships in December, a working assumption rather than a scheduled date.
Beyond Hegotá, the cluster organizes its work around five multi-fork research arcs:
- Fast finality, which aims to cut time to finality from minutes to seconds, with decoupled consensus the leading headliner candidate for the fork after Hegotá.
- Post-quantum security, anchored on a self-imposed target of a quantum-resistant Layer 1 across execution, consensus and data by December 2029.
- Privacy, starting in Hegotá with protocol support for trustless, censorship-resistant private transactions on Layer 1.
- State, covering a new trie, sustainable state growth and decentralized access to current and historical state.
- zkEVM, moving execution proofs from optional to mandatory so validators eventually verify proofs instead of re-executing every block.
Also Read: MARA Bitcoin Sale Unproven After $81M Reaches A Galaxy Digital Wallet
Conclusion: What Glamsterdam Means for Ethereum
Glamsterdam is less a single feature than a reorganization of Ethereum's base layer. It moves the builder market into the protocol, gives clients a map for parallel work and charges more honestly for permanent state.
The Sepolia launch went smoothly, but harder tests lie ahead. Potuz has pointed out that Hoodi hosts staking pools and home stakers, which makes it a far tougher fork than Sepolia.
A scheduling proposal from core developers sets the Hoodi fork at epoch 132,352, or 17:42:48 UTC on Oct. 26. No mainnet date has been agreed, and the foundation has said a separate announcement will cover it.
The open questions are economic as much as technical. Whether real builders show up, whether validators adopt the higher gas limit and whether demand fills the new space will decide what Glamsterdam actually delivers.
If it works, Ethereum's base layer will scale on its own terms, with rollups building on top rather than carrying the whole load.
Read Next: Evernorth Nasdaq Debut Moves To Oct. 12: What The Merger Delay Changes
