multithreading in *BEAST2

[Loïs Rancilhac]

Dear all,

I am currently inferring a species tree using *BEAST2 in BEAST2 and I would like to know if it is possible to run it on several threads ? I tried to use the -threads argument like with SNAPP, but the program only makes use of 1 core. I've read somewhere that *BEAST2 was not good at multithreading so maybe there is no point to try to run it on several cores ?

Also, I am thinking of using *BEAST2 to infer a species tree from a very big dataset (4000+ loci). Do some of you already tried to run it on such big datasets ? Does it sound feasible, particularly in term of computation time ?

Best regards,

Loïs Rancilhac.

[Remco Bouckaert]

Hi Loïs,

For many cases, StarBeast2 hardly benefits from multithreading, though you can try a few things suggested here: http://www.beast2.org/performance-suggestions/index.html

Cheers,

Remco

--
You received this message because you are subscribed to the Google Groups "beast-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to beast-users...@googlegroups.com.
To post to this group, send email to beast...@googlegroups.com.
Visit this group at https://groups.google.com/group/beast-users.
For more options, visit https://groups.google.com/d/optout.