Vitalik Buterin Calls for a Hard Ceiling on Ethereum Proving Costs

Vitalik Buterin wants a hard limit on proving costs

Vitalik Buterin, a co-founder of Ethereum, has called for a hard ceiling on the cost of producing proofs per unit of block gas. He made the request in a post on October 15, according to CryptoSlate.

In that post, Buterin congratulated ZKsync's Airbender team and named three priorities for formal verification: matching the mathematical representation of execution to Ethereum's specification, bounding prover cost relative to block gas, and securing the proof system itself.

A zero-knowledge proof is a way to show that a computation was done correctly without redoing all of it. Gas is Ethereum's unit for measuring and limiting the work a block can consume.

Key points

• Buterin's October 15 post set a proving-cost bound as one of three formal-verification priorities.
• EIP-7904 recommends no changes to Ethereum's compute gas costs, but its analysis does not specifically cover ZK-SNARK proof generation.
• ZKsync reported that a single NVIDIA RTX PRO 6000 Blackwell GPU proved 99% of 7,200 consecutive Ethereum mainnet blocks in under 10 seconds each, including witness preparation.
• Nethermind Research measurements found some operations are relatively expensive to prove per unit of gas.

Cost accounting sits apart from a speed leaderboard

CryptoSlate reports that the cost requirement goes beyond raw proving speed. Gas lets Ethereum account for work and cap a block's resource use. A useful proving bound would have to link that allowance to the resources needed to make a proof, including for workloads built to be difficult on purpose.

Security work was already underway. The Ethereum Foundation set out a security-focused phase in December 2025. The open question, per CryptoSlate, is whether the accounting that supports higher execution capacity can also keep the proving burden predictable.

What EIP-7904 does and does not cover

EIP-7904 is an Ethereum proposal that analyzes compute costs. CryptoSlate reports that, as of October 18, the proposal recommends no compute gas-cost changes and is an Informational EIP in Last Call, with a November 1 deadline.

The analysis targets 100 million gas per second on optimized native execution clients. It attributes its changed conclusion to optimizations enabled by Block-Level Access Lists, including parallel work that speeds up previously problematic operations. It specifies no changes to Ethereum's gas schedule.

The scope is limited: execution measured on a bare CPU. The document expressly says it does not specifically address ZK-SNARK proof generation. CPU execution and proof generation place different demands on hardware.

ZKsync reports single-GPU progress on proving

ZKsync's October 13 announcement described a single-GPU evaluation. One NVIDIA RTX PRO 6000 Blackwell proved 99% of 7,200 consecutive Ethereum mainnet blocks in under 10 seconds each, including witness preparation, and every sampled block finished in under 12 seconds, according to the company.

The announcement also described separate live ETHproofs configurations using two and four GPUs.

CryptoSlate notes that the sample records performance on those blocks. A maximum-cost guarantee would have to cover every permitted workload within its stated assumptions, including blocks deliberately assembled to be expensive to prove.

Research finds wide differences in proving cost per gas

CryptoSlate cites January 27 measurements by Nethermind Research's Lin Oshitani. The study found that cryptographic precompiles and several modular arithmetic or division operations were relatively expensive to prove per unit of gas. Costs varied by prover, and ZK-cycle counts did not translate into uniform elapsed proving time across operations.

The study estimated marginal proving time by varying operation counts while holding the surrounding execution context constant. It tested SP1 v5.2.3 with sp1-cluster and RISC0 v3.0.4, using reth v1.9.3 and four NVIDIA GeForce RTX 4090 GPUs, with RISC0 set to RISC0_KECCAK_PO2=15.

The measurements used fixed representative or best-effort worst-case inputs. They omitted block-header validation and accidentally omitted BLOCKHASH. A full investigation of argument-dependent costs remained future work.

Benchmarking framework ties proving throughput to block limits

CryptoSlate points to the Ethereum Foundation's January worst-case benchmarking framework, which relates required proving throughput to a chosen block gas limit and proving deadline. Under that approach, a difficult block still has to fit the intended hardware and time budget.

The framework also explains why an operation-by-operation estimate needs an end-to-end check. Proof generation has fixed overhead, aggregation can be nonlinear, and hardware configuration and security parameters affect performance. The set of execution clients and proof systems expected to take part also matters.

Where the burden lands

CryptoSlate reports that the engineering response determines who carries the cost. Improving an expensive circuit or execution path could cut proving work while leaving a transaction's execution gas units unchanged, so operators might need fewer resources for the same workload.

For an operator budgeting hardware, the distinction matters. A typical block may fit comfortably within a small setup, while a difficult legal workload determines the reserve capacity needed to keep producing proofs on time.

CryptoSlate notes that Nethermind's study also outlined weighted ZK-gas metering with a per-block ceiling, a system that would count proof-generation costs. The supplied material does not include further detail on that proposal.

What is confirmed

Buterin's October 15 post, his three formal-verification priorities, the status and scope of EIP-7904, the Ethereum Foundation's December 2025 security-focused phase, the ZKsync single-GPU results, and the Nethermind measurements are all reported in the supplied source.

What is still unclear

The source does not state whether Ethereum will adopt a hard proving-cost ceiling, what that limit would be, or when any change would take effect. The specific hardware and time budget that would define a worst-case proving envelope is not settled in the material provided. The Nethermind measurements carry stated limitations, including omitted block-header validation, an accidental omission of BLOCKHASH, and fixed inputs rather than a full survey of cost variation.

Why it matters for crypto users

Proving cost affects how much hardware independent prover operators need to keep up with Ethereum. If the hardest allowed blocks need far more proving power than typical ones, operators must reserve extra capacity. A clear limit would make those hardware and time budgets easier to plan, according to the source.

Sources

• CryptoSlate