Jia Huo
hjyyc0999@gmail.com
thank you very much for your suggestion.
now I first optimize the structure with G6+31(d,p) and then TZVPP, which only took one week.
can I ask another question:
(1)why I should add NZVAR=1? I check the document, which says NZVAR =M, M is usually 3N-6,
or 3N-5 for linear.
(2)I'm using the latest version of Firefly, should I still change
"1. You need to copy the MPI binding MPIBIND.NT-MPICH-SMP.DLL (from the BINDINGS subdirectory) and rename it as mpibind.dll in the root of the Firefly installation directory (replacing the default mpibind.dll that is provided). The mpich_smp.dll file must also be present in this same directory to work properly. See http://classic.chem.msu.su/gran/gamess/bindings.html for more details.
2. You need to rename the the following file pcgp2psm.dll to pcgp2p.dll (this will replace the default pcgp2p.dll that is located in the root of the Firefly installation directory). Look at http://classic.chem.msu.su/gran/gamess/p2p.html for more details."
best regards,
Jia
On Thu Feb 21 '13 7:17pm, Alex Granovsky wrote
----------------------------------------------
>Dear Jia Huo,
>>I found the total CPU utilization is less than 50ÐIs this normal for this kind of optimization.
>MP2 gradient is rather disk I/O intensive so that depending
>on the speed of your hard disk(s) the CPU usage may be not
>very high (although Firefly includes multiple tricks to make it as
>high as possible).
>>And also, even though I did DFT optimization before doing MP2 optimization, the structure was not optimized after more than 70 cycles. Is this system too big to do MP2 simulation or is there any problem in my input file?
>There is no problems with your input file except that optimization
>in DLCs would be much more efficient and require less cycles.
>Just add the following line to your input:
>
$CONTRL nzvar=1 $END
>Kind regards,
>Alex Granovsky
>
>
>On Sat Feb 2 '13 3:54am, Jia Huo wrote
>--------------------------------------
>>Dear All,
>>Recently, I tried to optimize a structure with 29 atoms including 4 Be using FF 8.0 with 2 cores.
>>I found the total CPU utilization is less than 50ÐIs this normal for this kind of optimization.
>>And also, even though I did DFT optimization before doing MP2 optimization, the structure was not optimized after more than 70 cycles. Is this system too big to do MP2 simulation or is there any problem in my input file?
>>the input file is as follow:
>>! PCGAMESS STYLE MP2 OPTIMIZATION CALCULATION OF BeO
>> $CONTRL SCFTYP=RHF MPLEVL=2 maxit=100 fstint=.true. runtyp=optimize $END
>> $CONTRL ICHARG=0 MULT=1 $END
>> $MP2 METHOD=1 $END
>> $p2p p2p=.t. dlb=.t. $end
>> $p2p mxbuf=2048 $end
>> $SYSTEM TIMLIM=1000000 MWORDS=250 $END
>> $BASIS EXTFIL=.T. GBASIS=TZVPP $END
>> $SCF DIRSCF=.True. $END
>> $GUESS GUESS=HUCKEL KDIAG=0 $END
>> $STATPT NSTEP=100 $END
>> $DATA
>>RHF/MP2/TZVPP/BeO
>>C1
>>Be 4.0 18.35750 33.20920 44.01900
>>Be 4.0 18.17340 31.06610 45.70640
>>Be 4.0 20.27050 32.85990 46.03440
>>Be 4.0 20.23010 31.17940 43.87550
>>O 8.0 19.25910 32.09230 44.91450
>>O 8.0 18.29830 32.84960 42.41050
>>O 8.0 17.95680 31.47480 47.29440
>>O 8.0 16.76050 33.25360 44.46200
>>O 8.0 21.87300 32.54610 45.80590
>>O 8.0 20.49860 29.65450 44.43920
>>O 8.0 20.18340 34.50950 45.95690
>>O 8.0 19.59930 31.00220 42.37340
>>O 8.0 19.94820 32.49600 47.59930
>>O 8.0 18.88170 34.75380 44.12550
>>O 8.0 18.65960 29.49980 45.74310
>>O 8.0 21.73630 31.82000 43.67610
>>O 8.0 16.65400 31.08830 45.07710
>>C 6.0 18.81490 31.84370 41.84400
>>C 6.0 18.85850 31.99920 48.01070
>>C 6.0 16.12870 32.19200 44.74020
>>C 6.0 22.37560 32.24640 44.68360
>>C 6.0 19.70510 29.03290 45.20390
>>C 6.0 19.61710 35.19440 45.05690
>>H 1.0 18.55520 31.68230 40.78600
>>H 1.0 18.68010 32.02440 49.09730
>>H 1.0 15.02950 32.22970 44.68210
>>H 1.0 23.46550 32.36550 44.57460
>>H 1.0 19.94910 27.98180 45.42460
>>H 1.0 19.78110 36.28330 45.08450
>> $END
>>
>>
>>i try to attach the out file.
>>many thanks,
>>Jia