Re: Parallel computing and JLink
- To: mathgroup at smc.vnet.net
- Subject: [mg94361] Re: Parallel computing and JLink
- From: Jens-Peer Kuska <kuska at informatik.uni-leipzig.de>
- Date: Thu, 11 Dec 2008 03:44:33 -0500 (EST)
- Organization: Uni Leipzig
- References: <ghg9bb$ho5$1@smc.vnet.net> <ghlm9k$jgt$1@smc.vnet.net> <gho2vm$t6o$1@smc.vnet.net>
- Reply-to: kuska at informatik.uni-leipzig.de
Hi, yes but the idea for "parallel Mathematica" it to use the wonderful Mathematica language and to do a parallel computation with it. When I have to switch to Java for the computation I loose the "wonderful Mathematica language" and than I can also use C/C++, FORTRAN or anything else. And if it should be fast, I mean really fast one has to use C/C++ anyway. Regards Jens mark mcclure wrote: > On Dec 9, 6:54 am, Jens-Peer Kuska <ku... at informatik.uni-leipzig.de> > wrote: > >> that's really amazing ! Your parallel JLink program is >> only 8.15 times slower that a parallel C++ program. On >> a dual core machine ... > > The original post was simply proof of concept, primarily in > response to another thread's claim that the new parallel > tools in V7 are incompatible with Mathematica's overall > design objectives. Clearly, other languages could be used, > as I even pointed out in the OP. > >> And the C++ program works without any parallel Mathematica >> stuff. > > Of course, the parallel programming could be done over in > the other language; the point is that it doesn't have to be. > >> I see that Java programs are still faster written than >> executed. > > Yes, as all serious users of Mathematica know, development > time is a crucial issue as well. Thank you for pointing > that out. > > Mark McClure >