Ethereum co-founder Vitalik Buterin has laid out how EIP-8037 could let the network process more than ten times its current execution capacity through the Glamsterdam and Hegotá upgrades, provided developers limit how much new data nodes must store permanently. The argument rests on an asymmetry that has troubled Ethereum for years: computation is cheap to scale, while state is not.
The mechanism is a repricing of the operations that create permanent data. Under the proposal, opening a new account would cost 197,900 gas instead of 36,600. Creating a storage slot would rise from 22,100 to 110,020 gas. Deploying a 24 KiB contract would require roughly 37.8 million gas, against 4.95 million today, with a separate state gas reservoir keeping large deployments possible despite the execution gas limit applied to individual transactions.
The numbers behind that repricing are the reason for it. Without a change, Buterin’s hypothetical tenfold increase in block capacity would push annual state growth from roughly 110 GiB to 1,100 GiB. Every full node would carry that data indefinitely. Charging more for state-heavy operations makes the cost of permanence visible to the contract that creates it, while leaving ordinary transfers and trades to benefit from the added capacity.
For ETH holders the transmission mechanism is fees, not capacity itself. Ethereum burns the base fee under EIP-1559, so greater usage can remove more coins from circulation. But more blockspace tends to push the base fee down, which means activity has to grow enough to offset the cheaper unit price. Whether the upgrade increases the burn depends on the balance between gas consumption, demand and the base fee at any moment. More capacity is a precondition for more fee revenue, not a guarantee of it.
Implementation carries real costs. Buterin acknowledged that contracts with fixed gas allowances, or with assumptions about how gas is refunded, could behave unexpectedly once the new accounting rules apply. Those assumptions are not exotic: a contract that forwards a set amount of gas to a call, or that counts on unused gas coming back to finish its own work, can revert or run dry when the schedule underneath it changes. Code that was correct on the day it shipped becomes a liability later.
Developers postponed Glamsterdam’s Hoodi activation while investigating exactly that, and security researchers went on to identify eight categories of assumptions that may fail. Two competing compatibility proposals have been floated. One would preserve the older gas rules for contracts already deployed; the other would run separate accounting systems to calculate execution costs.
Buterin said most core developers currently favour working directly with application teams to repair affected systems rather than adding further protocol complexity for the sake of backward compatibility. That is a bet that the long-run benefit of a cleaner gas schedule outweighs the cost of coordinating fixes across live applications, and it is the kind of judgement that is easier to defend in a design document than during an outage.
The work is already partly tested. Developers ran a 200 million gas limit on Sepolia on October 6. Hoodi remains the open question, and the answer will shape how much of the promised capacity arrives with Glamsterdam rather than later.
For the market, the proposal is a reminder that Ethereum’s scaling roadmap now runs through pricing policy as much as through new cryptography. Raising the price of state is a less quotable change than a throughput headline, but it is the condition on which the throughput depends.