https://bugs.kde.org/show_bug.cgi?id=428412
--- Comment #4 from Alexander Semke <alexander.se...@web.de> --- knights is using 1 core for the UI part and for the communication with the chess engines - this is fast enough. The actual computation of the board and of the moves is happening in the engines - this is where the parameters like the number of cores, the computation depth, etc. are relevant. At the moment knights can only control the depth and the memory of the hash tables for the XBoard protocol and the memory for the hash tables for the UCI protocol - this can be influenced by the user by selecting a different "difficulty". I don't see in the protocol specs how to influence the degree of parallelization from the client side, though. Knight would need to learn the different setting for the different chess engines like the smpmt-parameter for crafty (https://craftychess.com/documentation/craftydoc.html) and so on. But at the moment it's not planned to add this functionality to knights. So, if you want to let your favorite engine use multiple cores now you'll need to check and to modify its settings. -- You are receiving this mail because: You are watching all bug changes.
