[gmx-users] Folding at Home on PS3 == Gromacs on PS3?

Florian Haberl Florian.Haberl at chemie.uni-erlangen.de
Fri Aug 25 09:20:30 CEST 2006


Hi,

On Friday 25 August 2006 09:08, Erik Lindahl wrote:
> Hi,
>
> The Pande group will indeed be using Gromacs in F at H on the PS3, but
> at least initially it will be not be an accelerated version that
> makes good use of all the 8 SPE processors, but rather compiled. To
> them it's still useful due to the sheer number of machines they can
> get access to.
>
> However, we _are_ working on real Cell optimizations, but that's much
> harder. In contrast to SSE/Altivec/ia64 assembly it is not just a
> matter of instructions, but we need to come up with an entirely new
> way of handling the data, and programming direct memory access on the
> SPEs.

A good article about cell and scientific computing is:

http://portal.acm.org/citation.cfm?id=1128022.1128027

>
> Similar stuff is going in with graphics cards. Right now the
> neighborlists is the main obstacle, but we have good hopes this might
> improve in the next generation of GPUs.
>
> So, we're not keeping things secret, it's just that this type of
> hardware isn't particularly useful compared to clusters yet ;-)

Also clearspeed technology is interesting but not really usable nowadays.

>
> Cheers,
>
> Erik
>
> On Aug 24, 2006, at 9:38 PM, David Mobley wrote:
> > Sabuj,
> >
> > I suggest asking the Pande group directly. As far as I know they are
> > mostly not checking the list that regularly.
> >
> > Are you sure that it is F at H on the PS3? I had thought they just ported
> > GROMACS there.
> >
> > David
> >
> > On 8/24/06, Sabuj Pattanayek <sabuj.pattanayek at vanderbilt.edu> wrote:
> >> Hi all,
> >>
> >> I haven't used Folding at Home but I've heard that it uses Gromacs as
> >> its
> >> engine. So this probably means that someone has already ported it for
> >> the Cell proc. Any plans to eventually make this port available
> >> for d/l?
> >>
> >> Thanks,
> >> Sabuj Pattanayek
> >> _______________________________________________
> >> gmx-users mailing list    gmx-users at gromacs.org
> >> http://www.gromacs.org/mailman/listinfo/gmx-users
> >> Please don't post (un)subscribe requests to the list. Use the
> >> www interface or send it to gmx-users-request at gromacs.org.
> >> Can't post? Read http://www.gromacs.org/mailing_lists/users.php
> >
> > _______________________________________________
> > gmx-users mailing list    gmx-users at gromacs.org
> > http://www.gromacs.org/mailman/listinfo/gmx-users
> > Please don't post (un)subscribe requests to the list. Use the www
> > interface or send it to gmx-users-request at gromacs.org.
> > Can't post? Read http://www.gromacs.org/mailing_lists/users.php
>
> _______________________________________________
> gmx-users mailing list    gmx-users at gromacs.org
> http://www.gromacs.org/mailman/listinfo/gmx-users
> Please don't post (un)subscribe requests to the list. Use the
> www interface or send it to gmx-users-request at gromacs.org.
> Can't post? Read http://www.gromacs.org/mailing_lists/users.php


Greetings,

Florian

-- 
-------------------------------------------------------------------------------
 Florian Haberl                        
 Computer-Chemie-Centrum   
 Universitaet Erlangen/ Nuernberg
 Naegelsbachstr 25
 D-91052 Erlangen
 Telephone:  	+49(0) − 9131 − 85 26581
 Mailto: florian.haberl AT chemie.uni-erlangen.de
-------------------------------------------------------------------------------



More information about the gromacs.org_gmx-users mailing list