Gevulot Docs
Comment on page


The Gevulot network can be seen as having two distinct node types which together converge on network state: validators and provers. On a high level provers complete proving workloads and validators process transactions, verify proofs and order these into blocks.
The defining feature of the Gevulot network is the lack of re-execution. Proofs are generated by a user configurable subset of prover nodes after which a network defined subset of validators verify the proof. This configuration allows the network to parallelize compute both on an individual node level (running multiple provers/verifiers in parallel) as well as on a network level (different provers computing different proofs).
Note: We anticipate more node types in the future, such as non-validating full nodes, which just verify proofs and re-execute replicated state transitions.