Ethereum Co-founder Vitalik Buterin: Proofs to Replace Full Network Recalculation by 2030
According to CoinDesk, Ethereum co-founder Vitalik Buterin envisions that by 2030, Ethereum will rely more on cryptographic proofs and off-chain computation to process transactions, rather than having each node repeat the same calculations.
In this proposal, proofs will be primarily stored and verified on-chain, allowing users to independently confirm results without replaying all transactions, while enhancing privacy protection for payments, balances, and wallet activities. The verification path will shift from "downloading blocks and executing locally" to short proofs like SNARKs, with data availability relying on PeerDAS sampling instead of full transport.
This transformation is set to accelerate after the planned Hegota upgrade, which he describes as likely the last technical hard fork that developers from 2015 can still relate to as a "traditional" hard fork; thereafter, the main line will shift towards recursive STARKs, automated formal verification, optimized consensus, and quantum-resistant replacements. He also outlines engineering thresholds: the cost of generating proofs must be reduced, and safe coordination of parallel workflows is necessary; otherwise, scalability and privacy cannot coexist.
The 2030 roadmap also includes: authorization expanding from ordinary signatures to quantum-resistant signatures or zero-knowledge proofs, consensus moving towards fewer slots for finality, and block construction shifting from a single entity to multiple participants. The publicly discussed slot targets are about 4 to 8 seconds, with finality around 8 to 32 seconds. The decentralized network is designed not only for secure payments and censorship resistance but also to enhance performance through parallel storage, parallel computation, and privacy technologies.
In market mechanisms, buyers integrate the "verification layer" into the valuation of ETH, L2, and staking funds, while sellers continue to price L1 based on full node re-execution. Funding shifts from full network recalculation to proof-generating hardware, sampling bandwidth, and formal verification tools. Beneficiaries include infrastructure that can reliably produce proofs and perform sampling verification, while the old client paths that rely on replaying transactions to self-verify are under pressure. The event-driven shift is to write the 2030 model as "proofs on-chain, computation off-chain," rather than the next conventional fork list.
Source: Public Information
ABAB AI Insight
Vitalik divides the 2015 world computer into two layers: execution can move off the main chain, while verification must remain on the main chain. CoinDesk emphasizes that the billing unit has changed—not another fork name, but rather the network is no longer selling "every node computes it once" redundancy, but rather "one proof, multiple checks." Hegota is set as a boundary to stop core developers from using 2015's tool mindset to plan for 2028's protocols.
The capital path is the externalization of verification costs. Proof generation will create a new specialized supply, similar to today's rollup proof market and MEV construction market; the main chain will charge for data availability and finality rents. Resources will shift from full state node hard drives to proof clusters and sampling bandwidth. The motivation is to achieve both throughput and privacy: full network replay cannot keep balances and wallet activities confidential, while proofs combined with off-chain computation can.
Similar shifts include Bitcoin SPV, rollups pushing execution out of L1, and privacy pools removing transaction content from public ledgers. The industry is entering a control period: L2 continues to expand execution, while L1 defines who has the right to generate proofs accepted by the entire network and who cannot be excluded from the list. The transformation is incomplete, as proof costs and decoupling parallel workflows remain hard constraints.
Structural judgments belong to technological substitution. Repeated execution is replaced by short proofs, public balances are replaced by verifiable privacy, and single builders are replaced by multiple participants. The mechanism is: when verifying a proof is cheaper than replaying a transaction, full nodes are no longer the only secure model, and pricing power shifts from "who can synchronize the entire chain" to "who can prove computation cheaply and correctly."
ABAB News · Cognitive Laws
- Every node computing it once is not secure; it is wasteful.
- What should be stored on-chain is proof, not all intermediate steps.
- Privacy and scalability must coexist for a successful upgrade.