[Ns-developers] Status Parallelization

[Gustavo Carneiro]

2008/6/24 Cesar Marcondes <cesar at cs.ucla.edu>:

> Dear Hagen,
>
>   Sorry to send this unrelated message. But, I was wondering if you
> think is it possible (in the future) to have a CUDA implementation
> (using NVidia graphical cards) of ns-3?


I could be wrong, but I don't think GPUs are good for anything but simple
number crunching and processing arrays of numbers.  Network simulation, on
the other hand, has very little to do with crunching numbers and a lot to do
with conditional jumps and complex data structures.


>
>
> Best wishes,
> Cesar
>
> On Tue, Jun 24, 2008 at 11:21 AM, Hagen Paul Pfeifer <hagen at jauu.net>
> wrote:
> >
> > Hello NS3,
> >
> > short summary (although there is no great progress since last week, due
> to
> > required preparation time for a presentation and a disease): I rebased my
> > code and cleaned up all superfluous  parts towards a MPI clean
> > implementation. Yesterday I started to hook into the point to point
> channel
> > to tx/rx a Packet - based on federations. I hope that today in the
> evening
> > the code is ready to tx/rx a packet based on federation decisions (e.g.
> > process on a local node or peer node). So I will send today in the
> evening
> > an additional email with status information.
> >
> > Best regards, Hagen
> >
> >
> >
> >
>



-- 
Gustavo J. A. M. Carneiro
INESC Porto, Telecommunications and Multimedia Unit
"The universe is always one step beyond logic." -- Frank Herbert