On Mon, 7 Mar 2011, Ralf Wildenhues wrote:
FWIW I think the cumulative plots make sense when trying to answer the question âhow many packages have a speedup <= Xâ.Well, if I'd ask a cumulative question I'd probably rather ask how many packages have a speedup >= X. Small but significant difference.
Also, the nature of the packages. For example recursive/non-recursive. Total number of source lines (e.g. from 'sloccount'). Number of objects built per Makefile. Number of Makefiles. Number of libraries linked. Size (as from 'size' program) of final executable files.
Can you please make your hardware specs available? I've had two 16-way systems, and one could barely get more than a speedup of 2, while the other cucked away nicely at 10 or so.
Due to overlapping I/O, it is quite possible for builds to be super-linear. At least that is what I typically see here.
Bob -- Bob Friesenhahn bfriesen@xxxxxxxxxxxxxxxxxxx, http://www.simplesystems.org/users/bfriesen/ GraphicsMagick Maintainer, http://www.GraphicsMagick.org/
_______________________________________________ Autoconf mailing list Autoconf@xxxxxxx http://lists.gnu.org/mailman/listinfo/autoconf
