🚀 go-pugleaf

   jiu1@husc10.harvard.edu (Haibin Jiu) writes:

   > General question:  Does anyone have a compilation on Linda's performance?
   >                Esp. that of C-Linda?  On any system...

	   Numbers abound.  "Experience with Linda" (Bjornson et. al.,
YALEU/DCS/TR-866) has data on codes doing matrix multiplication, LU
decomposition, parallel sequence comparison, fluid dynamics and more.
Linda tech reports can be obtained from Chris Hatchell
(hatchell@cs.yale.edu).

   > Specific question:  We seem to have been unable to run more than 4 or 5
   >                C-Linda processes on a LAN of Sun SPARCstations?  Even
   >                the simplest programs -- some taken from a paper by
   >                the guy who started Linda and some other dude on parallel
   >                programming -- lock up if more than 4 Suns are involved.
   >                Any comments?


	Without knowing more about your problem, there is not much I
can say other than to note that Linda programs routinely run on much
larger collections of machines.

   	BTW, with respect to a comment from an earlier posting, SCA's
next release (v2.4.7, in use now at Yale) supports multiple processes
per node.


David