[Wien] Benchmark-Compilation options

Clazo Clazo clazo36 at web.de
Fri Jun 16 10:49:37 CEST 2006 

Dear Prof. Blaha,


Thank you for your answers. I was finally able to reach a time very close to the one reported by Martin Hilgeman

ifort 9.1 + MKL (8.02) + libgoto_ITANIUM2-r1.02.so

ALTIX4700 Itanium2 Madison 9M Prozessoren (1.66GHz/6MB L3 Cache)

test_case running time: 124.5 Sec

For those interested, my compilation options are:

     O   Compiler options:        -O3 -tpp2 -ftz -ip -fno-alias -cm -w -FR -DALTIX -DINTEL_VML
     L   Linker Flags:            -L../SRC_lib -Vaxlib -L/opt/intel/mkl/8.0.2/lib/64 -L/fibus/fs1/16/vt4cl/LIB/GotoBLAS
     P   Preprocessor flags       '-DParallel'
     R   R_LIB (LAPACK+BLAS):     -lgoto_ITANIUM2-r1.02 -lmkl_lapack64 -lmkl -lvml


The problem before was that the GOTO libraries were not being linking properly. I had to install to patches which are available in the GOTO Blas web-site.  For that, one has to unpack the patches (.diff files) in the directory where the GOTO library is an issue the command:

patch -p1 < patch_name.diff

After that the compilation runs OK and one can generate the *.so version of the library according to the instructions in the install.txt file.

Best Regards,

Cesar Lazo
Technische Universität Hamburg-Harburg









> -----Ursprüngliche Nachricht-----
> Von: A Mailing list for WIEN2k users <wien at zeus.theochem.tuwien.ac.at>
> Gesendet: 15.06.06 08:08:02
> An: A Mailing list for WIEN2k users <wien at zeus.theochem.tuwien.ac.at>
> Betreff: Re: [Wien] Benchmark-Compilation options


> Just noticed, that your timing does not seem to be cortrect. Check 
> cputim.c; Some machines report the cputime in different units and you 
> should divide by 1024 instead of 100.
> 
> I've also heared from an Bull Itanium system and they gave very similar 
> timings as the Altix. So I guess these values should be reachable. Since 
> your "Hamilt-time" is ok (besides the factor 10), it is mainly the 
> gotoblas, which makes the difference. Also your report that you get the 
> same timings with/without goto tells me, that you are NOT using the 
> gotolib (or not a good version). I guess, the goto-libraries were changed 
> with version 1.0 and one has to compile them himself, so maybe older 
> versions (precompiled) were more efficient .... ???
> 
> Always remember: The hamilt time is purely f90 + svml-lib code, hns is 
> mixed f90 + blas, and diag is purely blas-code. So compare the partial 
> times.
> 
> > Thank you for your confirmation. I have tried really hard to reproduce the 122 sec posted in the Wien2K benchmark page, unfortunately without success.
> > 
> > I get exactly the same as you. After x lapw1 -c in the test_case:
> > 
> > TIME HAMILT (CPU)  =   133.6, HNS =   222.4, DIAG =  1331.3
> >        NUMBER OF K-POINTS:           1
> >    ===> TOTAL CPU       TIME:   1710.1 (INIT =     21.3 + K-POINTS =   1688.8)
> >    > SUM OF WALL CLOCK TIMES:    167.4 (INIT =      2.2 + K-POINTS =    165.2)
> >       Maximum WALL clock time:    167.726181983948     
> >       Maximum CPU time:           1710.76000000000 
> > 
> > Interestingly, I get the same with and without GOTO libraries. Here are my compilation options corresponding to the times listed above:
> > 
> >      O   Compiler options:        -O3 -tpp2 -ftz -ip -fno-alias -cm -w -FR -DALTIX -DINTEL_VML
> >      L   Linker Flags:            -L../SRC_lib -Vaxlib -L/opt/intel/mkl/8.0.2/lib/64
> >      P   Preprocessor flags       '-DParallel'
> >      R   R_LIB (LAPACK+BLAS):     -lmkl_lapack64 -lmkl -lvml
> > 
> > An here are the times of Martin Hilgeman on a similar machine (as posted in a recent email):
> > 
> > 
> >        TIME HAMILT (CPU)  =    11.2, HNS =    15.8, DIAG =    94.0
> >    ===> TOTAL CPU       TIME:    123.2 (INIT =      2.1 + K-POINTS =
> > 121.1)
> >    > SUM OF WALL CLOCK TIMES:    123.6 (INIT =      2.2 + K-POINTS =
> > 121.4)
> >       Maximum WALL clock time:    123.825130939484
> >       Maximum CPU time:           123.211914062500
> > 
> > I carried out the test calculations with Wien2K version VERSION_05.6 while Martin used version WIEN2k_06.1. I do not believe that the different versions are the cause of such difference in running times (aprox. 45 sec.); however, I will check it. Maybe the problem has to to with the differences in CPU times. I will investigate more about this. Please, let me know if you manage to reduce the running time of the test_case.
> > 
> > Regards
> > 
> > 
> > Cesar Lazo
> > Technische Universität Hamburg-Harburg
> > 
> > 
> > 
> > 
> > 
> > 
> > > -----Ursprüngliche Nachricht-----
> > > Von: A Mailing list for WIEN2k users <wien at zeus.theochem.tuwien.ac.at>
> > > Gesendet: 14.06.06 18:44:28
> > > An: "A Mailing list for WIEN2k users" <wien at zeus.theochem.tuwien.ac.at>
> > > Betreff: Re: [Wien] Benchmark-Compilation options
> > 
> > 
> > > For information;
> > > 
> > > I compiled W2K as Cesar Clazo did'it  (ifort9.0 +mkl8.0, libgoto_itanium2_64p-r1.00)
> > > 
> > > Results for the benchmark in a ALTIX  Itanium2 @ 1.5 GHz with 4 MB in  L3  are ;
> > > 
> > >     ********************************************************
> > > 
> > >        NUMBER OF K-POINTS:           1
> > >    ===> TOTAL CPU      
> > > TIME:   1734.1 (INIT =     21.3 +
> > > K-POINTS =   1712.9)
> > >    > SUM OF WALL CLOCK TIMES:    169.8
> > > (INIT =      2.2 + K-POINTS
> > > =    167.6)
> > >       Maximum WALL clock time:    169.997846841812     
> > >       Maximum CPU time:           1734.78000000000 
> > > 
> > > 
> > > best regards
> > > 
> > > Dr. Emilio Orgaz
> > > Departamento de Física y Química Teórica
> > > Facultad de Química, UNAM
> > > Emilio.Orgaz at gmail.com
> > > 
> > > 
> > > 
> > > On 6/9/06, Clazo Clazo <clazo36 at web.de> wrote:
> > > Dear all,
> > > 
> > > I
> > > compiled the wien code  VERSION_05.6:
> > > 20.10.2005  in a SGI Altix 4700 system with Itanium2 Madison
> > > 9M Prozessoren (1.66GHz/6MB L3 Cache) , ifort version 9.0 + mkl8.0 +
> > > libgoto_ITANIUM2-r1.00.a,  and I carried out the standard
> > > test_case provided in the wien2k web site in order to see how well the
> > > tunning of the compiler and libraries was. I obtained a running time of
> > > 160 Sec. The running time for a similar setup reported in the wien2k
> > > Benchmarks site is:
> > > 
> > > Itanium2(1.6GHz,SGI Altix 3700)  122 sec    ifort9.0 +mkl8.0, libgoto_itanium2_64p-r1.00
> > > 
> > > Thus,
> > > faster than the time I got. The compiling options  (most of
> > > them are the default options recommended by the siteconfig script) are
> > > the following:
> > > 
> > >      O   Compiler
> > > options:        -O3 -tpp2 -ftz
> > > -ip -fno-alias -cm -w -FR -DALTIX -DINTEL_VML
> > >     
> > > L   Linker
> > > Flags:            -L../SRC_lib
> > > -Vaxlib
> > >      P   Preprocessor flags       '-DParallel'
> > >      R   R_LIB (LAPACK+BLAS):     -L/fibus/fs1/16/vt4cl/LIB/GotoBLAS/libgoto_ITANIUM2-r1.00.a -lscs -L/opt/intel/mkl/8.0.2/lib/64/ -lvml
> > > 
> > > I downloaded and compiled the last version of GOTO library from the web.
> > > 
> > > 
> > > Has
> > > somebody having a similar system been able to reproduce the running
> > > time posted in the wien2k Benchmark page (122 sec)? If so, could you
> > > post the options used to compile the code?
> > > 
> > > The running time I
> > > get (160 sec) is already good, but I think, if the code is able to run
> > > faster in the computational setup available, I should try to tune the
> > > compiler and libraries in order to get the best possible performance.
> > > 
> > > Thank you in advance,
> > > 
> > > 
> > > Cesar Lazo
> > > Technische Universität Hamburg-Harburg
> > > 
> > > 
> > > 
> > > 
> > > 
> > > 
> > > _____________________________________________________________________
> > > 
> > > Der WEB.DE SmartSurfer hilft bis zu 70% Ihrer Onlinekosten zu sparen!
> > > http://smartsurfer.web.de/?mc=100071&distributionid=000000000071
> > > 
> > > 
> > > _______________________________________________
> > > Wien mailing list
> > > Wien at zeus.theochem.tuwien.ac.at
> > > 
> > > http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> > > 
> > > 
> > > 
> > > -- 
> > > Dr. Emilio Orgaz
> > > Departamento de Física y Química Teórica
> > > Facultad de Química, UNAM
> > > Emilio.Orgaz at gmail.com
> > > Tel.  5622-3776    
> > > Fax. 5622-3521
> > > 
> > > -----------------------------------------------------------------
> > > _______________________________________________
> > > Wien mailing list
> > > Wien at zeus.theochem.tuwien.ac.at
> > > http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> > > 
> > 
> > 
> > ____________________________________________________________________________
> > ____________________________________________________________________________
> > WM Spa&szlig; mit Style! Coole Spr&uuml;che, Outfits & Co. rund um Deine
> > Lieblingsspieler als kostenlose eCard versenden! http://www.sendaplayer.net
> > 
> > _______________________________________________
> > Wien mailing list
> > Wien at zeus.theochem.tuwien.ac.at
> > http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> > 
> 
> 
>                                       P.Blaha
> --------------------------------------------------------------------------
> Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
> Phone: +43-1-58801-15671             FAX: +43-1-58801-15698
> Email: blaha at theochem.tuwien.ac.at    WWW: http://info.tuwien.ac.at/theochem/
> --------------------------------------------------------------------------
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien


_____________________________________________________________________
Der WEB.DE SmartSurfer hilft bis zu 70% Ihrer Onlinekosten zu sparen!
http://smartsurfer.web.de/?mc=100071&distributionid=000000000071

More information about the Wien mailing list