[Date Index] [Thread Index] [Author Index]

Re: Message passing in mathematica?

• To: mathgroup at smc.vnet.net
• Subject: [mg99589] Re: Message passing in mathematica?
• From: Jens-Peer Kuska <kuska at informatik.uni-leipzig.de>
• Date: Fri, 8 May 2009 00:19:16 -0400 (EDT)
• Organization: Uni Leipzig
• References: <gtudpg$j5r$1@smc.vnet.net>
• Reply-to: kuska at informatik.uni-leipzig.de

Hi,

http://daugerresearch.com/pooch/mathematica.shtml

?

Regards
   Jens

gidiko wrote:
> Hello Mathematica Users,
> 
>>From what I understand Mathematica does not currently offer "seamless"
> support for the message passing paradigm in its parallel offerings (I
> mean apart from Link*[] primitives, however even then send/receive-
> like primitives should be programmed by the user).
> 
> Is there some package supplying convenient message passing API atop
> MathLink protocol? (start a "cluster" of workers each given its ID,
> send/receive primitives, perhaps collective operations (gather/
> scatter)...
> 
> In some occasions the virtual shared memory paradigm currently offered
> in Mathematica is not convenient when porting parallel algorithms laid
> down in e.g. MPI.
> 
> Thanks,
> Giorgos
> 
>