Hello folks, Going from a sequential application to a parallel one is far from trivial, and this would need in-depth knowledge of the code architecture. It’s my understanding that there are successive phases involved in LP, and that's not good news for parallelization.
And there’s Guile in the loop: do we have an estimation of the fraction of the time spend in Guile when using LP for typical scores? This amount of time will probably be incompressible. JM Le 4 juin 2014 à 11:00:25, Mike Solomon <m...@mikesolomon.org> a écrit : > On Jun 4, 2014, at 10:51 AM, Orm Finnendahl > <orm.finnend...@hfmdk-frankfurt.de> wrote: > >> Hi, >> >> rendering a score on an i7-2640M with the option "-djob-count=2" >> shows 100% cpu load only on one thread of one of the two cores while >> rendering. >> >> Is there some way or a recommended technique of setting up lilypond to >> better distribute the load? Or does lilypond have to get compiled with >> special compile options in order to take full advantage of >> multithreading? >> >> -- >> Orm > > Just a shout out to Orm that I’m working on a research project now for > multi-threading in musical engraving via openmp - lemme know if that is of > interest and I can send you some stuff. > > Cheers, > MS > _______________________________________________ > lilypond-user mailing list > lilypond-user@gnu.org > https://lists.gnu.org/mailman/listinfo/lilypond-user _______________________________________________ lilypond-user mailing list lilypond-user@gnu.org https://lists.gnu.org/mailman/listinfo/lilypond-user
