On 2012-03-27, Volker Braun <vbraun.n...@gmail.com> wrote: > The current state is that we build static atlas with and without threads, > and the shared library without threads only. And our module_list.py links > with the single-threaded atlas shared library only. So Sage will use the > single-threaded version if you build atlas yourself. > > If you use the os-provided atlas library (which can be multi-threaded yet > not have the "pt" in the name), you get parallel atlas. For example, on > Fedora I'm using the system atlas and get > > sage: sage: a = random_matrix(RDF,1000) > sage: time a*a > 1000 x 1000 dense matrix over Real Double Field > Time: CPU 1.08 s, Wall: 0.26 s > > And if anybody wonders why there is a single-threaded atlas at all: People > often specifically WANT a single-threaded blas, and implement their own > parallelism on top of that.
Actually, it's strange that one wants to build own parallel things on top a single-threaded blas. It seems to go against the original philosophy of blas, which is that the vendor (think Cray :)) provides super-duper parallel blas... In my area, it's usually the other way around: people rely on parallel blas/lapack to speed things up. So I too vote for multithreaded ATLAS to be on in Sage by default. Dima -- To post to this group, send an email to sage-devel@googlegroups.com To unsubscribe from this group, send an email to sage-devel+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URL: http://www.sagemath.org
