[gmx-users] GROMACS with SGI native compliers

Nathan Moore nmoore at physics.umn.edu
Thu Aug 4 18:40:08 CEST 2005


Dear Ken,

Here's what I mean with regard to machine specific compilation.  I ran
this on an old SGI O2 workstation at the U of Minnesota.

first determine the machine properties with the uname command,

11:35am|nmoore at i7> uname -a

IRIX64 i7 6.5 04070317 IP30

So it is an SGI, running 64bit IRIX 6.5 with name "i7" and processor type
"IP30".  When you compile source, give Ofast the machine type,

11:35am|nmoore at i7> cc hello.c -LIST:options -Ofast=IP30

Nathan





> With best intentions, I'd suggest you start by reading the Man pages for
> the SGI IRIX compiler that's installed on your Origin.  These man pages
> generally have the best description of how to compile a binary for a
> specific MIPS processor (I think optimization goes all the way down to the
> chipset - ie not all R12k processors are optimized with the same flag).
>
> Also, if you're doing a bunch of PME stuff you might want to either (1)
> replace FFTW with the SGI-specific LAPACK routine (I think has name
> similar to "complib.sgimath" or else (2) talk to the FFTW developers to
> see if there's a version of FFTW that's optimized for your R14k
> architecture.
>
> regards,
>
> NT
>
>
>
>> Hello All,
>>
>> We've been successfully running GROMACS on an SGI supercomputer using
>> the gcc complier. However, we'd really like to realize the speed
>> increases that should be possible using the MIPSpro 7.4 IRIX native
>> compliers.
>>
>> This what we're running on:
>>
>> SGI Origin 3800, 128 processor R14000 500 MHz
>> SGI Message Passing Toolkit 1.6 (this is SGI's tuned MPI libraries)
>> FFTW version 2.1.3 (single-precision, shared libraries, MPI-enabled)
>> GCC 3.3 (SGI supplied) and SGI MIPSpro 7.4 compilers
>> Latest GROMACS config flags: --enable-shared --enable-mpi CC=gcc
>> CFLAGS=-L/usr/lib32 CPPFLAGS=-I/usr/local/include
>> LDFLAGS=-L/usr/local/lib
>>
>> I'd love to hear from anyone who's running GROMACS on a similar
>> system... especially the kind individual at SGI that offered to help
>> the last time I posted w/SGI issues, whose e-mail I've apparently
>> misplaced/lost.
>>
>> thanks in advance,
>>
>> Ken
>>
>> K.S. Rotondi Ph.D.
>> Research Fellow
>> The Gierasch Laboratory
>> Department of Biochemistry and Molecular Biology
>> University of Massachusetts-Amherst
>> Phone: 413-545-1250
>> Fax;      413-545-3291
>>
>> _______________________________________________
>> 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.
>>
>
>




More information about the gromacs.org_gmx-users mailing list