Around 88% of Ethereum blocks are currently built off-chain, through trust-based relays sitting between validators and builders.
That arrangement was never supposed to be permanent. Glamsterdam is the upgrade that tries to end it.
Core developers describe it as the largest protocol change since The Merge.
The Name and the Structure
Glamsterdam is a portmanteau. Gloas is the consensus-layer component, Amsterdam is the execution-layer component, and they ship together as one network upgrade.
Ethereum names forks by pairing a star with the city that hosted the most recent Devconnect. It is the direct successor to Fusaka, which shipped in December 2025.
Two headline proposals anchor it, with a supporting cluster of gas-repricing and state-cost changes around them.
EIP-7732 Moves Block Building Into the Protocol
Enshrined proposer-builder separation replaces the off-chain relay system with in-protocol bids, commitments and payments.
Today that role is filled by MEV-Boost, which works but requires validators to trust relays they do not control. Enshrining it removes that trust assumption.
There is a second effect that matters more for throughput. The roadmap describes ePBS expanding the propagation and validation window from roughly 2 seconds to about 9 seconds.
That is the constraint Ethereum has been fighting. A 12-second slot leaves very little room to validate a large block without pushing home operators off the network. More time means bigger blocks become safe.
EIP-7928 Makes Parallel Execution Possible
Block-Level Access Lists attach a record to every block mapping each account and storage slot the block touches, along with post-execution values.
The organizing principle is address rather than transaction. Each account appears once, with every transaction that touched it referenced inside.
Ethereum currently discovers transaction dependencies during execution, which is why transactions cannot safely run in parallel. Declaring them upfront removes that problem.
Three things follow. Nodes can perform parallel disk reads and validate independent transactions simultaneously. State root computation can run in parallel. And syncing nodes can update from recorded results rather than replaying every transaction.
That last one is quietly significant. Executionless sync would remove the slow peer-negotiation phase that snap sync currently needs.
| Proposal | Layer | What it does |
| EIP-7732 (ePBS) | Consensus | Moves block building in-protocol, widens validation window |
| EIP-7928 (BALs) | Execution | Declares state access upfront, enables parallel execution |
| EIP-8159 | Networking | Method for clients to exchange BALs |
| Gas and state repricing | Execution | Supporting cost adjustments |
The Scaling Claim, Carefully Stated
Researcher Toni Wahrstätter, one of the EIP-7928 authors, has framed the two proposals together as roughly a 10x scaling improvement from baseline with no hardware upgrades required.
Reporting also indicates Glamsterdam is expected to ship at a 200 million block gas limit before pushing higher, and that ePBS gives validators up to four times more execution time.
Separately, one analysis puts the combined effect at roughly tripling layer 1 capacity.
Those figures do not agree, and they should not be treated as interchangeable. They measure different things under different assumptions, and none of them is a mainnet result.
BALs do not guarantee a fixed transaction speed. What they do is make raising the gas limit further technically safe, which is the mechanism by which capacity actually increases.
The Date Is Not Locked
Ethereum’s own roadmap uses an H2 2026 window. Core developers have been working toward Q4 2026, and no mainnet date has been confirmed.
The upgrade reached final devnet testing in June 2026. Recent developer calls have focused on devnet-5 with devnet-6 scope finalization named as the next step.
Public testnets including Sepolia and Hoodi will activate before mainnet. Anyone tracking this should watch testnet activation rather than roadmap language.
One EIP in the supporting cluster remains in draft, so the scope is real but not frozen.
What Was Cut, and Why
FOCIL, fork-choice-enforced inclusion lists for censorship resistance, was scoped out of Glamsterdam and moved to a later upgrade named Hegotá.
The stated reason is delivery discipline. ePBS touches nearly every layer of the block production path, and adding more to that scope was judged too risky.
That is a healthier signal than it sounds. Ethereum’s upgrade history includes repeated scope reductions to protect timelines, and the alternative is indefinite delay.
What Holders and Operators Need to Do
ETH holders need to do nothing. Balances, addresses and private keys are unaffected.
Stakers and node operators must update both consensus-layer and execution-layer clients before mainnet activation. That is the entire action item, and missing it means running an incompatible node.
Optimisus has covered this pattern through several Ethereum cycles, including when the Pectra upgrade completed on mainnet.
Why This Upgrade Is Strategically Different
Recent Ethereum upgrades optimized for rollups. Blobs, data availability, layer 2 economics.
Glamsterdam points back at layer 1. The stated goal is making Ethereum itself a better venue for stablecoin settlement and on-chain real-world assets, rather than pushing that activity to rollups.
That shift matters for anyone holding ETH on a value-accrual thesis, because base-layer activity and rollup activity route fees very differently.
It also lands while the network is debating validator economics, a fight Optimisus covered when a proposal appeared to cut staking rewards toward zero.
Capacity and issuance are being reconsidered at the same time. How those two land together will define Ethereum’s economics more than either does alone.
Solana is running its own consensus replacement on a similar timeline, covered in the Alpenglow validator rollout. Two of the largest smart contract networks are rebuilding their base layers in the same window.
Sources
- Ethereum Foundation, Glamsterdam roadmap page — https://ethereum.org/roadmap/glamsterdam/
- Ethereum Improvement Proposals, EIP-7928 Block-Level Access Lists — https://eips.ethereum.org/EIPS/eip-7928
- Chainstack, Ethereum’s Glamsterdam upgrade, what changes for infrastructure — https://chainstack.com/ethereum-glamsterdam-upgrade/
- Everstake, Ethereum Glamsterdam upgrade overview and EIPs explained — https://everstake.one/resources/blog/ethereum-glamsterdam-upgrade-explained
- Datawallet, Ethereum Glamsterdam upgrade and EIPs explained — https://www.datawallet.com/crypto/ethereum-glamsterdam-upgrade-explained
- ETH Daily, Block-Level Access Lists explained — https://ethdaily.io/block-level-access-lists-explained
Optimisus covers crypto and technology news for readers who want the detail behind the headline.
