Ethereum co-founder Vitalik Buterin is backing EIP-8037, a proposal for the network’s upcoming Glamsterdam upgrade that would sharply raise the gas cost of creating new accounts, storage slots, and smart contracts. The idea is to let Ethereum process far more transactions per block without forcing node operators to store an unsustainable amount of new permanent data, but the tradeoff is real: the change could break existing contracts and bridges that hardcode gas limits for operations that are about to get a lot pricier.

Buterin laid out his reasoning in an article he shared on Monday, arguing that Ethereum’s two main scaling constraints can’t keep growing at the same pace forever. Execution capacity, meaning how many transactions a block can process, is comparatively easy to scale up. State, the permanent record of every account, contract, and storage slot that a full node has to keep on disk, is not. Letting both grow in lockstep eventually prices ordinary operators out of running a node at all.
“The goal of EIP-8037 is to decouple Ethereum’s scaling limits: enable a much higher amount of execution (which is easy to scale), but only a little bit more state size growth (which is hard to scale),” Buterin wrote.
He backed that up with a specific estimate. Ethereum adds roughly 110 GiB of state per year today. Raise the block gas limit tenfold without repricing the operations that create new state, Buterin wrote, and that figure could climb to around 1,100 GiB a year. That kind of growth curve pushes up disk requirements, slows initial sync times for anyone trying to stand up a new node, and makes it progressively harder for everyday operators to keep running Ethereum at home rather than leaving it to a shrinking pool of well-resourced providers.
What gets more expensive
EIP-8037’s fix is to make the specific actions that create new state dramatically more expensive, while leaving other execution cheap enough that the overall gas limit can be raised later without the same tradeoff. Using Buterin’s own figures, the cost of creating a brand-new account would jump from 36,600 to 197,900 gas units, more than five times the current price. A new storage slot would rise from 22,100 to 110,020 gas units. Deploying a full 24 KiB smart contract would climb from about 4.95 million gas to nearly 37.8 million, roughly a 7.6x increase.
One detail will land on ordinary users, not just developers: the new account-creation charge also applies when a plain ETH transfer sends funds to an address that doesn’t exist on-chain yet. That’s a routine pattern for wallets, exchanges, and payroll-style payouts, so the fee increase doesn’t stay confined to people deploying contracts.

Buterin didn’t gloss over the proposal’s biggest weakness. Raising gas costs by this much risks breaking live applications, particularly bridges and other contracts that hardcode gas limits for specific operations. A contract built on the assumption that creating an account costs around 36,600 gas will simply fail once that number jumps to nearly 200,000. Developers will need to audit their contracts and transaction configurations for exactly this kind of hardcoded assumption before Glamsterdam ships, and Buterin called that work out directly rather than burying it as a footnote.
That tradeoff puts EIP-8037 in a familiar spot for Ethereum governance: a change that’s hard to argue against in principle, since almost nobody wants state bloat to quietly price ordinary users out of running nodes, but one that shifts real migration work onto the builders the network needs to keep around. It’s the same tension that showed up when Solana’s community debated retiring durable nonces after the Drift exploit: a fix that’s correct on paper still has to clear the bar of not breaking what’s already running in production.
Why this matters beyond Ethereum mainnet
EIP-8037 is also a signal about where execution-layer competition is heading across the wider Ethereum ecosystem. Layer 2s and alternative execution environments have spent the past year racing to prove they can process more transactions more cheaply, and state growth is exactly the constraint that limits how far that race can go without compromising who can actually run a node. MegaETH’s pivot away from a pure high-throughput DeFi chain toward an AI compute marketplace happened partly because raw throughput alone wasn’t enough to sustain a chain’s economics long term. Buterin’s proposal is the mainnet version of the same underlying problem: speed is cheap to advertise, but somebody has to pay for the state it leaves behind.
Nothing here is finalized yet. EIP-8037 is a proposal, not a shipped feature, and Ethereum’s core developers still have to agree on the exact numbers before Glamsterdam locks its scope. But Buterin putting his name behind a specific pricing table, rather than a vague call for “better gas accounting” down the road, is usually the point where a proposal starts moving from forum debate toward an actual client implementation.
The bottom line
Buterin is picking the less popular but more honest tradeoff here. Repricing state creation hits developers and bridge operators in the short term, and some of them are going to have a rough few months retrofitting gas assumptions that have sat unchanged for years. But the alternative, letting state grow in lockstep with execution capacity, has a worse ending: a network only well-funded operators can afford to run, which is the opposite of what Ethereum is supposed to be. A nearly 7.6x price increase on contract deployment sounds dramatic until you remember that state is permanent and gas is not. Charging more upfront for something that sits on every node forever is the kind of pricing problem that’s far easier to fix before Glamsterdam ships than after.



