Re: Cluster computing drawbacks
- From: glen herrmannsfeldt <gah@xxxxxxxxxxxxxxxx>
- Date: Tue, 26 Jul 2005 10:48:02 -0700
Nick Maclaren wrote:
(snip)
Unless the application is a complete crock, it needs NO rewriting to run under a MPI spawning program (which are widespread), a batch scheduler or both. Just Do It. Converting to use MPI communication is harder, but still easier than converting to use SMP communication.
Most people's experience is that it is EASIER than converting a serial program to use SMP communication. Seriously. Converting to use SMP is one of the foulest tasks that you can imagine, and is enought to make you wake up in the middle of the night screaming and dripping with sweat.
I thought people just ran MPI code on the SMP machines...
Ones that I would be interested in working on are PDE solvers,
and the communications might be a little faster on SMP, at least with not so many processors.
-- glen
.
- References:
- Cluster computing drawbacks
- From: Emidio S.
- Re: Cluster computing drawbacks
- From: Randy
- Re: Cluster computing drawbacks
- From: already5chosen
- Re: Cluster computing drawbacks
- From: Randy
- Re: Cluster computing drawbacks
- From: already5chosen
- Re: Cluster computing drawbacks
- From: Ketil Malde
- Re: Cluster computing drawbacks
- From: Nick Maclaren
- Re: Cluster computing drawbacks
- From: Ketil Malde
- Re: Cluster computing drawbacks
- From: Nick Maclaren
- Cluster computing drawbacks
- Prev by Date: Re: Cluster computing drawbacks
- Next by Date: Re: Cluster computing drawbacks
- Previous by thread: Re: Cluster computing drawbacks
- Next by thread: Re: Cluster computing drawbacks
- Index(es):