This paper describes the new features of the distributed theorem prover Peers-mcd.d for equational logic. They include parallel-search plans that combine distributed search (i.e., the search space is subdivided among the parallel processes) and multi-search (i.e., the processes employ different search plans). Experiments comparing the behavior of different control strategies on the Moufang identities are reported, including instances of super-linear speed-up due to parallel search.
|Titolo:||Combination of distributed search and multi-search in Peers-mcd.d|
|Data di pubblicazione:||2001|
|Appare nelle tipologie:||04.01 Contributo in atti di convegno|