Ethereum's first publicly accessible testnet for its upcoming Glamsterdam hard fork goes live on August 20, 2026. Named Platåberget (Swedish for "plateau mountain"), the network gives developers worldwide their first hands-on opportunity to test changes widely described by protocol researchers as the most significant protocol restructuring since the 2022 Merge. The testnet launch comes with a direct warning: any tool with a hardcoded maximum gas limit will stop working.

---

What Is Glamsterdam?

Glamsterdam combines two codenames: Gloas, the consensus-layer component named after a star, and Amsterdam, the execution-layer component referencing the 2022 Devconnect host city. It succeeds the Fusaka upgrade from December 2025. The fork bundles roughly ten Ethereum Improvement Proposals (EIPs) that collectively raise the block gas limit from approximately 60 million to 200 million, embed block-builder competition directly into the protocol, introduce a separate accounting system for state growth, and raise the maximum deployable smart contract size from 24 KiB to 64 KiB.

---

Market Context

ETH was trading near $1,900 on August 17, down roughly 35% year-to-date from a 2025 peak near $5,000. Average transaction fees sit at about 0.047 gwei, or less than $0.20 per standard transfer, near multi-year lows and down roughly 95% from 2024 and 2025 peaks. Glamsterdam is designed to push throughput toward 10,000 transactions per second, compared to roughly 1,000 today, according to protocol analysts at Phemex.

---

The Most Consequential Changes

EIP-7732 removes the third-party relay infrastructure that currently processes about 88% of Ethereum blocks. Under the existing system, a separate ecosystem of trusted relays coordinates block building outside the core protocol through a setup called MEV-Boost. EIP-7732 moves this coordination on-chain, requiring builders to submit cryptographic bids that proposers can accept without trusting an intermediary. Validators will take on a new duty inside a Payload Timeliness Committee. Ethereum researcher Luca Zanolini summarized the shift (via X): "EIP-7732 enshrines proposer-builder separation into Ethereum's consensus protocol, removing the need for the trusted relays that today's MEV pipeline depends on." The change also extends the data propagation window from approximately 2 seconds to approximately 9 seconds, with practical implications for block timing and validator behavior.

EIP-7928 introduces Block-Level Access Lists, which track every account and storage slot a block touches alongside post-execution state values. These lists are what make the 200 million gas ceiling feasible. Without them, tripling the gas limit would produce unsustainable database growth. With them, clients can process higher throughput while maintaining coherent state. Block-Level Access Lists also lay the groundwork for two additional capabilities: parallel transaction execution and stateless client reconstruction, both significant for Ethereum's long-term scalability.

EIP-8037 is the most impactful breaking change for developers and wallets, a characterisation supported directly by the protocol research underlying this upgrade. It creates a separate "state gas" meter that charges for every byte added to Ethereum's state, covering actions like account creation, contract deployment, and storage writes. A plain ETH transfer to a new address, currently priced at a flat 21,000 gas, will now carry an additional state gas charge on top of execution gas. Wallets, gas estimators, and indexers built around the 21,000 assumption will need to be updated before mainnet. The Ethereum Foundation blog stated directly: "Any tool that relies on a hardcapped maximum gas limit (think wallets, indexers and gas estimators) will break and needs to be updated."

EIP-7954 lifts the deployed contract size ceiling from 24 KiB to 64 KiB and raises maximum initcode from 48 KiB to 128 KiB. Initcode is the bytecode executed at contract deployment and represents a distinct constraint from deployed contract size; both increases matter to developers deploying complex contracts via constructor logic. Complex DeFi protocols have hit these ceilings for years, forcing developers to split logic across multiple proxy contracts. The increases remove that constraint outright.

EIP-8061 raises validator exit and consolidation churn limits, a change with direct relevance to institutional staking operations. Benjamin Thalman, Staff Protocol Analyst at Figment, noted that Glamsterdam "materially enhances how quickly validators can exit large staked ETH positions," a meaningful improvement for institutions managing large validator sets that require more agile position management.

---

Regional Stakes Are High

For developers in South Asia and Africa, the timeline is tight. The Asia-Pacific region processed $2.36 trillion in on-chain value in 2025, up 69% year-over-year, and accounts for 36.4% of global Web3 developer activity. India alone accounts for roughly 17% of all new Web3 developers globally, the highest share of any single country. Devcon 8, the flagship Ethereum developer conference, is confirmed for Mumbai in Q4 2026, the same window as Glamsterdam's targeted mainnet launch. Local infrastructure teams at Polygon, Devfolio, ETHMumbai, and university blockchain clubs will need to audit gas estimation logic and contract deployment workflows before mainnet pressure arrives.

In Africa, Nigeria leads the continent with more than 45% crypto adoption among internet users, according to Web3 Economy Statistics and CoinLaw, and Kenya also ranks among the world's highest adoption markets by proportion of internet users. ETHSafari 2026, Africa's main Ethereum community gathering, runs in Nairobi and Kilifi from September 4 to 12, just weeks before Sepolia and Hoodi (Ethereum's longer-lived public testnets) receive the Glamsterdam fork. That overlap gives African developers a concrete window to engage with the changes before mainnet. The state gas model carries particular relevance in both regions: applications handling remittances, peer-to-peer payments, and on-chain identity for unbanked users will face changed cost structures any time they send ETH to a new address.

EIP-7708, which makes ETH transfers emit ERC-20-style Transfer logs, will simplify monitoring for fintech teams that already handle ERC-20 event pipelines and want consistent tooling for ETH flows.

---

What Developers Must Do Now

The Ethereum Foundation's warning is direct: tools with hardcoded gas assumptions will break. The following categories require action before mainnet:

• Wallets: Update gas estimation logic to account for the new state gas meter introduced by EIP-8037. ETH transfers to new addresses will cost more than 21,000 gas.
• Gas estimators: Any tool that assumes a fixed maximum gas limit or flat transfer cost must be updated to handle the dual execution gas and state gas accounting model.
• Indexers: Indexers that parse transaction receipts or gas usage based on current limits will need schema and logic updates to handle the 200 million gas ceiling and the new state gas fields.
• Smart contract deployments: Contracts near the current 24 KiB size limit should be tested against the new 64 KiB ceiling. Constructor logic should be validated against the new 128 KiB initcode limit.
• Validator software and node operators: Validators should review the Payload Timeliness Committee duties introduced by EIP-7732 and account for the extended data propagation window. Node operators should confirm client compatibility; all major clients are supported, including Lighthouse, Prysm, Geth, Nethermind, and Reth.

---

What Comes Next

After Platåberget, the Glamsterdam fork is scheduled for Sepolia and Hoodi in September 2026, followed by mainnet in Q4 2026. The upgrade following Glamsterdam, codenamed Hegotá, is slated for the first half of 2027 and will include two features deferred from this cycle: Fork-Choice Inclusion Lists (FOCIL) and a proposal to reduce slot times from 12 to 6 seconds. FOCIL was deferred because of the complexity of its interaction with enshrined proposer-builder separation at mainnet scale. The 6-second slot proposal was deferred because ZK-proving technology is not yet mature enough to support it reliably.

Developers can access Platåberget through the developer portal at plataberget.dev, the block explorer at dora.plataberget.ethpandaops.io, and the testnet ETH faucet at faucet.plataberget.ethpandaops.io. Support is available via the Ethereum R&D Discord and through the consensus-specs, execution-specs, and execution-apis repositories. All major consensus and execution clients, including Lighthouse, Prysm, Geth, Nethermind, and Reth, are supported.