Techniques that let a party outsource a computation and receive a proof that the result is correct, allowing efficient verification without redoing the work. Proof systems such as SNARKs and STARKs make verification exponentially cheaper than re-execution, enabling trust-minimised delegation at scale.
Semantic Classification
Content
- Verifiable computation allows a client to delegate a computation to an untrusted server and then check, using a succinct proof, that the returned result is correct without re-executing the computation. Proof systems such as succinct non-interactive arguments make verification far cheaper than the original work.
- The technique underpins scaling approaches that move computation off-chain while proving its integrity, as well as privacy-preserving applications where a result can be verified without revealing the inputs. Hardware-based attestation through trusted execution environments offers a complementary trust model.