Ethereum's Glamsterdam upgrade has officially gone live on Sepolia testnet on October 6 at 13:53:36 UTC, marking one of its final public testing stages before planned mainnet activation. According to Ethereum community contributor Pooja Ranjan, the upgrade activated on Sepolia at 1:53 p.m. UTC, putting changes to block production, parallel processing and gas pricing through a public network test. The activation follows weeks of preparation after developers scheduled the Sepolia fork, with node operators required to update both their execution and consensus clients before the fork. As per CoinDesk, the change was activated on Tuesday, October 6, and blocks were produced under the higher limit by Thursday, October 8.
Ethereum developers released a last-minute Prysm update version 7.2.1 on October 5 to ensure validators automatically produce blocks with the intended 200 million gas limit during the Glamsterdam upgrade's Sepolia test. According to Prysm's release notes, the update includes Sepolia's new gas limit schedule so validators automatically propose blocks using the 200 million setting once Glamsterdam activates. Prysm's previous release 7.2.0 supported the Sepolia fork but was completed before the 200 million setting was added to Sepolia's configuration, meaning validators running the older version would have defaulted to 60 million gas after activation unless operators manually changed the number by hand.
Early testing on Sepolia has revealed that blocks are using significantly less than their full capacity, with 26% to 46% utilization of the new 200 million gas limit. According to CoinDesk's review of over 25 consecutive test blocks, each block used roughly 52 million to 92 million gas, demonstrating that the threefold increase in processing budget does not translate directly into three times as many transactions. This conservative utilization pattern indicates the network can handle substantially more activity before reaching full capacity, with none of the test blocks running close to full utilization where the larger budget would put the most strain on network computers. The upgrade changes how work is divided between specialists assembling transactions and validators publishing blocks, giving validators more time to check transaction calculations. As per 99Bitcoins, the Sepolia blocks sampled so far used less than half their allowance, making the result a demonstration of headroom rather than proof that the network is already processing near-200-million-gas blocks at full load.
Curve Finance has identified significant potential benefits for complex DeFi transactions under the Glamsterdam upgrade, citing mainnet checks that show leveraged borrowing consuming 1.84 million gas compared to just 125,000 gas for direct stablecoin swaps. According to Curve's analysis, this represents a 14.7 times increase in gas usage for borrowing operations, highlighting the congestion challenges faced by DeFi users. The protocol expects the added capacity to particularly benefit borrowers, arbitrage traders and liquidators during periods of heavy network demand, as cheaper execution could make smaller arbitrage trades economically viable and improve liquidation mechanisms like LLAMMA. Curve cautioned that while the upgrade should help complex transactions, revised state charges under EIP-8037 and EIP-8038 could increase gas requirements for some actions, requiring contract and wallet updates before mainnet activation.
The upgrade introduces EIP-7732 with enshrined proposer builder separation (ePBS) and EIP-7928 with block level access lists (BALs), fundamentally changing how Ethereum blocks are built and processed. EIP-7732 separates block proposers and builders directly into Ethereum's consensus protocol, allowing validators to check execution payloads without performing the same role as the builder. The proposal separates consensus validation from execution validation, giving validators additional time to check execution payloads. EIP-7928 introduces block level access lists, recording accounts and storage locations accessed during a block along with relevant post-transaction state changes, enabling clients to read data from disk in parallel and execute transactions concurrently. These changes link Ethereum's higher-capacity ambitions to a broader scaling and Layer-2 capacity outlook.
