Vitalik Buterin says GKR could boost speed of zero-knowledge
-bXqJXTWy.png "Vitalik Buterin says GKR could boost speed of zero-knowledge")
Ethereum co-founder Vitalik Buterin has spotlighted the GKR (Goldreich–Kahan–Rothblum) protocol, which he says could greatly improve the speed and cost-efficiency of zero-knowledge (ZK) proofs.
Buterin explains that GKR enables faster verification of large computations by removing the need to repeatedly confirm every step of the process. Instead, it checks only the inputs and outputs, mathematically verifying that each transition between layers is correct. This dramatically reduces the amount of work a prover has to perform.
In theory, the method can cut proving overheads by up to 100x, and by roughly 10x in early real-world tests. He referenced current ZK benchmarks such as ZK-EVM systems that already prove Ethereum mainnet activity in real time using only ~50 consumer GPUs, as well as setups capable of verifying millions of Poseidon hashes per second on standard laptops.
According to Ethereum’s mastermind, GKR works best for tasks that are built in layers and repeated many times – a structure common in both blockchain hashing (such as Poseidon2) and AI inference workloads. Because of this, GKR could be applied not only to ZK rollups but also to machine learning proving systems.
According to Ethereum’s mastermind, GKR works best for tasks that are built in layers and repeated many times – a structure common in both blockchain hashing (such as Poseidon2) and AI inference workloads. Because of this, GKR could be applied not only to ZK rollups but also to machine learning proving systems.
He added that GKR-based systems avoid heavy use of commitment schemes like Merkle trees or KZG, which are normally required to verify every internal computation. Instead, they convert each round into a smaller polynomial proof step, reducing complexity and speeding up verification.
Vitalik noted that GKR by itself is not private – it only ensures correctness. To enable full zero-knowledge (privacy-preserving) capabilities, it must be combined with SNARKs or STARKs.
If adopted more broadly, GKR-based proving could lower the cost of ZK rollups, bring faster block verification, support verifiable AI inference, and make lightweight clients more practical on consumer hardware. Buterin says that as further optimizations reduce overhead, zero-knowledge verification may become efficient enough to serve as a standard layer for both blockchain scalability and AI-driven applications.
The full technical breakdown, titled "GKR Protocol," was published on Buterin’s blog.
Vitalik noted that GKR by itself is not private – it only ensures correctness. To enable full zero-knowledge (privacy-preserving) capabilities, it must be combined with SNARKs or STARKs.
If adopted more broadly, GKR-based proving could lower the cost of ZK rollups, bring faster block verification, support verifiable AI inference, and make lightweight clients more practical on consumer hardware. Buterin says that as further optimizations reduce overhead, zero-knowledge verification may become efficient enough to serve as a standard layer for both blockchain scalability and AI-driven applications.
The full technical breakdown, titled "GKR Protocol," was published on Buterin’s blog.
Author
Sebile Fane cut her teeth in blockchain by building tiny NFT experiments with friends in her living room, long before the buzzwords took hold. She’s driven by a curiosity for the human stories behind smart contracts — whether it’s a small-town artist minting her first token or a DAO voting on climate grants — and weaves technical insight with genuine empathy.
