[gmx-users] new i9 processor

[Mark Abraham]

Hi,

There's no real issue there. If anything, it's simpler. With the dual
socket system it would often make no sense to extend an OpenMP region
across the sockets. That's still true if there's internal structure that
makes some cores further from others (I dunno if true for Skylake-X). You'd
still want to use GROMACS in the same general way.

A 14-core node would make more complex structure harder to set up and use,
because 7 is prime. For example, a multi-sim setup using two simulations
each of two ranks on such a node would not have a natural way to split over
four ranks.

Mark

On Mon, Oct 16, 2017 at 11:59 AM Harry Mark Greenblatt <
harry.greenblatt at weizmann.ac.il> wrote:

> BS”D
>
> Dear All,
>
>   The new Skylake-X CPU’s have very high core counts, and could, in
> theory, take the place of a two socket 16 core Xeon system.  One could
> create a machine as follows:
>
> 1 x Core i9 7960X with 16 cores
> 2 x Geforce 1080 graphic cards,
>
> and run a simulation with 2 MPI ranks, coupling 8 cores to 1 gpu per rank.
>
> Would there be any technical problems arising from the single socket
> configuration, as opposed to a dual socket Xeon setup?  Any issues if a 14
> core processor was used (7 cores per rank, i.e., an odd number)?
>
> Thanks
>
> Harry
>
>
>
>
> --------------------------------------------------------------------
> Harry M. Greenblatt
> Associate Staff Scientist
> Dept of Structural Biology           harry.greenblatt at weizmann.ac.il
> <../../owa/redir.aspx?C=QQgUExlE8Ueu2zs5OGxuL5gubHf97c8IyXxHOfOIqyzCgIQtXppXx1YBYaN5yrHbaDn2xAb8moU.&URL=mailto%3aharry.greenblatt%
> 40weizmann.ac.il>
> Weizmann Institute of Science        Phone:  972-8-934-6340
> 234 Herzl St.                        Facsimile:   972-8-934-3361
> Rehovot, 7610001
> Israel
>
> --
> Gromacs Users mailing list
>
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-request at gromacs.org.