The large-scale transformation of the gas model as part of the Glamsterdam upgrade carries hidden risks for a significant portion of the ecosystem. My analysis shows that this is not about cosmetic changes, but a fundamental shift in fee mechanics that will affect critically important infrastructure.
The Ethereum Foundation (EF) team published a warning on August 17 that the upcoming update could take a number of software solutions offline. At particular risk are crypto wallets, blockchain data indexers, and services that estimate fee sizes. Developers are strongly advised to stress-test their systems on the public Plataberget test network.
The core of the problem: EIP-8037 and a new gas dimension
The root of the issues lies in the Ethereum improvement proposal EIP-8037. This initiative introduces a separate state-gas dimension, which will apply to operations that create new state on the blockchain. The key change: a standard ETH transfer to an existing address will retain the usual cost of 21,000 gas units, but sending coins to a new, previously unused address will result in an additional fee charge.
This creates a serious problem for applications that rely on outdated logic. The EF emphasizes that tools with a "hardcoded maximum gas limit" are especially vulnerable—they will simply stop working without an update. In addition, smart contracts that assume 21,000 gas units covers any ETH transfer, as well as code that calculates fees based solely on the amount of gas consumed, will need to be revised.
Timeline and prospects
According to monitoring services, the Plataberget testnet was launched on August 13, 2026, with Glamsterdam activation scheduled for August 20. This will be followed by a month-long feedback and client software update phase, after which testing will begin on the Sepolia and Hoodi networks. This means developers have limited time to adapt their products.
It is worth noting that in August, Ethereum co-founder Vitalik Buterin named quantum protection and artificial intelligence as priority areas for network development. However, the current warning shows that even without exotic technologies, basic protocol changes require maximum focus from the entire community.
My verdict: Glamsterdam is not just another hard fork, but a serious test of the ecosystem's maturity. Projects that ignore the EF warning risk losing users and incurring reputational costs. I recommend teams start auditing their solutions now, without waiting for the update to activate on the mainnet.