应《网络安全法》要求，自2017年10月1日起，未进行实名认证将不得使用互联网跟帖服务。为保障您的帐号能够正常使用，请尽快对帐号进行手机号验证，感谢您的理解与支持！

24小时热门版块排行榜    

小木虫论坛-学术科研互动平台 » 计算模拟区 » 第一性原理 » Vasp&MedeA » 如何提高vasp.5的速度
北京石油化工学院2026年研究生招生接收调剂公告
51/1返回列表
查看: 1093  |  回复: 2
只看楼主 @他人 存档 新回复提醒 (忽略) 收藏    在APP中查看
当前只显示满足指定条件的回帖，点击这里查看本话题的所有回帖

huangyc

木虫 (正式写手)

[求助] 如何提高vasp.5的速度

差不多配置的机器，5.2的速度比4.6的还要慢。5.2在执行时top一下发现，总有核处于S（睡眠)状态。




我的makefile
.SUFFIXES: .inc .f .f90 .F
#-----------------------------------------------------------------------
# Makefile for Intel Fortran compiler for Pentium/Athlon/Opteron
# bases systems
# we recommend this makefile for both Intel as well as AMD systems
# for AMD based systems appropriate BLAS and fftw libraries are
# however mandatory (whereas they are optional for Intel platforms)
#
# The makefile was tested only under Linux on Intel and AMD platforms
# the following compiler versions have been tested:
#  - ifc.7.1  works stable somewhat slow but reliably
#  - ifc.8.1  fails to compile the code properly
#  - ifc.9.1  recommended (both for 32 and 64 bit)
#  - ifc.10.1 partially recommended (both for 32 and 64 bit)
#             tested build 20080312 Package ID: l_fc_p_10.1.015
#             the gamma only mpi version can not be compiles
#             using ifc.10.1
#  - ifc.11.1 strongly recommended (we use this to compile vasp)
#             Build 20090630 Package ID: l_cprof_p_11.1.046
#
# it might be required to change some of library pathes, since
# LINUX installation vary a lot
# Hence check ***ALL*** options in this makefile very carefully
#-----------------------------------------------------------------------
#
# BLAS must be installed on the machine
# there are several options:
# 1) very slow but works:
#   retrieve the lapackage from ftp.netlib.org
#   and compile the blas routines (BLAS/SRC directory)
#   please use g77 or f77 for the compilation. When I tried to
#   use pgf77 or pgf90 for BLAS, VASP hang up when calling
#   ZHEEV  (however this was with lapack 1.1 now I use lapack 2.0)
# 2) more desirable: get an optimized BLAS
#
# the two most reliable packages around are presently:
# 2a) Intels own optimised BLAS (PIII, P4, PD, PC2, Itanium)
#     http://developer.intel.com/software/products/mkl/
#   this is really excellent, if you use Intel CPU's
#
# 2b) probably fastest SSE2 (4 GFlops on P4, 2.53 GHz, 16 GFlops PD,
#     around 30 GFlops on Quad core)
#   Kazushige Goto's BLAS
#   http://www.cs.utexas.edu/users/kgoto/signup_first.html
#   http://www.tacc.utexas.edu/resources/software/
#
#-----------------------------------------------------------------------

# all CPP processed fortran files have the extension .f90
SUFFIX=.f90

#-----------------------------------------------------------------------
# fortran compiler and linker
#-----------------------------------------------------------------------
FC=ifort
# fortran linker
FCL=$(FC)


#-----------------------------------------------------------------------
# whereis CPP ?? (I need CPP, can't use gcc with proper options)
# that's the location of gcc for SUSE 5.3
#
#  CPP_   =  /usr/lib/gcc-lib/i486-linux/2.7.2/cpp -P -C
#
# that's probably the right line for some Red Hat distribution:
#
#  CPP_   =  /usr/lib/gcc-lib/i386-redhat-linux/2.7.2.3/cpp -P -C
#
#  SUSE X.X, maybe some Red Hat distributions:

CPP_ =  ./preprocess <$*.F | /usr/bin/cpp -P -C -traditional >$*$(SUFFIX)

# this release should be fpp clean
# we now recommend fpp as preprocessor
# if this fails go back to cpp
CPP_=fpp -f_com=no -free -w0 $*.F $*$(SUFFIX)

#-----------------------------------------------------------------------
# possible options for CPP:
# NGXhalf             charge density   reduced in X direction
# wNGXhalf            gamma point only reduced in X direction
# avoidalloc          avoid ALLOCATE if possible
# PGF90               work around some for some PGF90 / IFC bugs
# CACHE_SIZE          1000 for PII,PIII, 5000 for Athlon, 8000-12000 P4, PD
# RPROMU_DGEMV        use DGEMV instead of DGEMM in RPRO (depends on used BLAS)
# RACCMU_DGEMV        use DGEMV instead of DGEMM in #-----------------------------------------------------------------------

CPP     = $(CPP_)  -DHOST=\"LinuxIFC\" \
          -DCACHE_SIZE=12000 -DPGF90 -Davoidalloc -DNGXhalf \
          -DRPROMU_DGEMV  -DRACCMU_DGEMV

#-----------------------------------------------------------------------
# general fortran flags  (there must a trailing blank on this line)
# byterecl is strictly required for ifc, since otherwise
# the WAVECAR file becomes huge
#-----------------------------------------------------------------------

#FFLAGS =  -FR -lowercase -assume byterecl
FFLAGS= -I/public/software/intel/Compiler/11.1.064/mkl/include/fftw -FR -lower_case
#FREE=-FR

#-----------------------------------------------------------------------
# optimization
# we have tested whether higher optimisation improves performance
# -axK  SSE1 optimization,  but also generate code executable on all mach.
#       xK improves performance somewhat on XP, and a is required in order
#       to run the code on older Athlons as well
# -xW   SSE2 optimization
# -axW  SSE2 optimization,  but also generate code executable on all mach.
# -tpp6 P3 optimization
# -tpp7 P4 optimization
#-----------------------------------------------------------------------

# ifc.9.1, ifc.10.1 recommended
OFLAG=-O1 -xSSE4.2

OFLAG_HIGH = $(OFLAG)
OBJ_HIGH =
OBJ_NOOPT =
DEBUG  = -FR -O0
INLINE = $(OFLAG)

#-----------------------------------------------------------------------
# the following lines specify the position of BLAS  and LAPACK
# VASP works fastest with the libgoto library
# so that's what we recommend
#-----------------------------------------------------------------------

RACC (depends on used BLAS)
# tbdyn                 MD package of Tomas  Bucko
# mkl.10.0
# set -DRPROMU_DGEMV  -DRACCMU_DGEMV in the CPP lines
#BLAS=-L/opt/intel/mkl100/lib/em64t -lmkl -lpthread
BLAS=-L/public/software/intel/Compiler/11.1.064/mkl/lib/em64t -lmkl_intel_lp64 -lmkl_blacs_lp64 -lmkl_intel_thread -lmkl_core -liomp5 -lpthread

# even faster for VASP Kazushige Goto's BLAS
# http://www.cs.utexas.edu/users/kgoto/signup_first.html
# parallel goto version requires sometimes -libverbs
#BLAS=  /opt/libs/libgoto/libgoto.so

# LAPACK, simplest use vasp.5.lib/lapack_double
#LAPACK= ../vasp.5.lib/lapack_double.o
LAPACK=-L/public/software/intel/Compiler/11.1.064/mkl/lib/em64t -lmkl_intel_lp64 -lmkl_blacs_lp64 -lmkl_intel_thread -lmkl_core -liomp5 -lpthread

# use the mkl Intel lapack
#LAPACK= -lmkl_lapack

#-----------------------------------------------------------------------

#LIB  = -L../vasp.5.lib -ldmy \
     ../vasp.5.lib/linpack_double.o $(LAPACK) \
     $(BLAS)

# options for linking, nothing is required (usually)
LINK    =

#-----------------------------------------------------------------------
# fft libraries:
# VASP.5.2 can use fftw.3.1.X (http://www.fftw.org)
# since this version is faster on P4 machines, we recommend to use it
#-----------------------------------------------------------------------

FFT3D   = fft3dfurth.o fft3dlib.o

# alternatively: fftw.3.1.X is slighly faster and should be used if available
#FFT3D   = fftw3d.o fft3dlib.o   /opt/libs/fftw-3.1.2/lib/libfftw3.a


#=======================================================================
# MPI section, uncomment the following lines until
#    general  rules and compile lines
# presently we recommend OPENMPI, since it seems to offer better
# performance than lam or mpich
#
# !!! Please do not send me any queries on how to install MPI, I will
# certainly not answer them !!!!
#=======================================================================
#-----------------------------------------------------------------------
# fortran linker for mpi
#-----------------------------------------------------------------------

FC=mpif90
FCL=$(FC)

#-----------------------------------------------------------------------
# additional options for CPP in parallel version (see also above):
# NGZhalf               charge density   reduced in Z direction
# wNGZhalf              gamma point only reduced in Z direction
# scaLAPACK             use scaLAPACK (usually slower on 100 Mbit Net)
# avoidalloc          avoid ALLOCATE if possible
# PGF90               work around some for some PGF90 / IFC bugs
# CACHE_SIZE          1000 for PII,PIII, 5000 for Athlon, 8000-12000 P4, PD
# RPROMU_DGEMV        use DGEMV instead of DGEMM in RPRO (depends on used BLAS)
# RACCMU_DGEMV        use DGEMV instead of DGEMM in RACC (depends on used BLAS)
# tbdyn                 MD package of Tomas  Bucko
#-----------------------------------------------------------------------

#-----------------------------------------------------------------------

CPP    = $(CPP_) -DMPI  -DHOST=\"LinuxIFC\" -DIFC \
     -DCACHE_SIZE=4000 -DPGF90 -Davoidalloc -DNGZhalf \
     -DMPI_BLOCK=8000 \
     -DRPROMU_DGEMV  -DRACCMU_DGEMV

#-----------------------------------------------------------------------
# location of SCALAPACK
# if you do not use SCALAPACK simply leave that section commented out
#-----------------------------------------------------------------------

#BLACS=$(HOME)/archives/SCALAPACK/BLACS/
#SCA_=$(HOME)/archives/SCALAPACK/SCALAPACK

#SCA= $(SCA_)/libscalapack.a  \
# $(BLACS)/LIB/blacsF77init_MPI-LINUX-0.a $(BLACS)/LIB/blacs_MPI-LINUX-0.a $(BLACS)/LIB/blacsF77init_MPI-LINUX-0.a

SCA=
#-----------------------------------------------------------------------
# libraries for mpi
#-----------------------------------------------------------------------

LIB     = -L../vasp.5.lib -ldmy  \
      ../vasp.5.lib/linpack_double.o\
       -L/public/software/intel/Compiler/11.1.064/mkl/lib/em64t\
      -lmkl_intel_lp64 -lmkl_scalapack_lp64 -lmkl_core -limf -lguide -lm \
      $(BLAS)
#     $(SCA) $(LAPACK)


# FFT: fftmpi.o with fft3dlib of Juergen Furthmueller
#FFT3D   = fftmpi.o fftmpi_map.o fft3dfurth.o fft3dlib.o

# alternatively: fftw.3.1.X is slighly faster and should be used if available
FFT3D   = fftmpiw.o fftmpi_map.o  fftw3d.o fft3dlib.o /public/software/intel/Compiler/11.1.064/mkl/interfaces/fftw3xf/libfftw3xf_intel.a


#-----------------------------------------------------------------------
# general rules and compile lines
#-----------------------------------------------------------------------
BASIC=   symmetry.o symlib.o   lattlib.o  random.o


SOURCE=  base.o     mpi.o      smart_allocate.o      xml.o  \
         constant.o jacobi.o   main_mpi.o  scala.o   \
         asa.o      lattice.o  poscar.o   ini.o  mgrid.o  xclib.o  vdw_nl.o  xclib_grad.o \
         radial.o   pseudo.o   gridq.o     ebs.o  \
         mkpoints.o wave.o     wave_mpi.o  wave_high.o  \
         $(BASIC)   nonl.o     nonlr.o    nonl_high.o dfast.o    choleski2.o \
         mix.o      hamil.o    xcgrad.o   xcspin.o    potex1.o   potex2.o  \
         constrmag.o cl_shift.o relativistic.o LDApU.o \
         paw_base.o metagga.o  egrad.o    pawsym.o   pawfock.o  pawlhf.o   rhfatm.o  paw.o   \
         mkpoints_full.o       charge.o   Lebedev-Laikov.o  stockholder.o dipol.o    pot.o \
         dos.o      elf.o      tet.o      tetweight.o hamil_rot.o \
         steep.o    dimer.o  dynmat.o  neb.o  lanczos.o  instanton.o  sd.o  cg.o  qm.o  lbfgs.o  bfgs.o  fire.o   opt.o  \
         chain.o    dyna.o     sphpro.o    us.o  core_rel.o \
         aedens.o   wavpre.o   wavpre_noio.o broyden.o \
         dynbr.o    rmm-diis.o reader.o   writer.o   tutor.o xml_writer.o \
         brent.o    stufak.o   fileio.o   opergrid.o stepver.o  \
         chgloc.o   fast_aug.o fock.o     mkpoints_change.o sym_grad.o \
         mymath.o   internals.o dynconstr.o dimer_heyden.o dvvtrajectory.o vdwforcefield.o \
         hamil_high.o nmr.o    pead.o     mlwf.o     subrot.o   subrot_scf.o \
         force.o    pwlhf.o  gw_model.o optreal.o   davidson.o  david_inner.o \
         electron.o rot.o  electron_all.o shm.o    pardens.o  paircorrection.o \
         optics.o   constr_cell_relax.o   stm.o    finite_diff.o elpol.o    \
         hamil_lr.o rmm-diis_lr.o  subrot_cluster.o subrot_lr.o \
         lr_helper.o hamil_lrf.o   elinear_response.o ilinear_response.o \
         linear_optics.o linear_response.o   \
         setlocalpp.o  wannier.o electron_OEP.o electron_lhf.o twoelectron4o.o \
         ratpol.o screened_2e.o wave_cacher.o chi_base.o wpot.o local_field.o \
         ump2.o bse_te.o bse.o acfdt.o chi.o sydmat.o dmft.o \
         rmm-diis_mlr.o  linear_response_NMR.o
INC=

vasp: $(SOURCE) $(FFT3D) $(INC) main.o
        rm -f vasp
        $(FCL) -o vasp main.o  $(SOURCE)   $(FFT3D) $(LIB) $(LINK)
makeparam: $(SOURCE) $(FFT3D) makeparam.o main.F $(INC)
        $(FCL) -o makeparam  $(LINK) makeparam.o $(SOURCE) $(FFT3D) $(LIB)
zgemmtest: zgemmtest.o base.o random.o $(INC)
        $(FCL) -o zgemmtest $(LINK) zgemmtest.o random.o base.o $(LIB)
dgemmtest: dgemmtest.o base.o random.o $(INC)
        $(FCL) -o dgemmtest $(LINK) dgemmtest.o random.o base.o $(LIB)
ffttest: base.o smart_allocate.o mpi.o mgrid.o random.o ffttest.o $(FFT3D) $(INC)
        $(FCL) -o ffttest $(LINK) ffttest.o mpi.o mgrid.o random.o smart_allocate.o base.o $(FFT3D) $(LIB)
kpoints: $(SOURCE) $(FFT3D) makekpoints.o main.F $(INC)
        $(FCL) -o kpoints $(LINK) makekpoints.o $(SOURCE) $(FFT3D) $(LIB)

clean:
        -rm -f *.g *.f *.o *.L *.mod ; touch *.F

main.o: main$(SUFFIX)
        $(FC) $(FFLAGS)$(DEBUG)  $(INCS) -c main$(SUFFIX)
xcgrad.o: xcgrad$(SUFFIX)
        $(FC) $(FFLAGS) $(INLINE)  $(INCS) -c xcgrad$(SUFFIX)
xcspin.o: xcspin$(SUFFIX)
        $(FC) $(FFLAGS) $(INLINE)  $(INCS) -c xcspin$(SUFFIX)

makeparam.o: makeparam$(SUFFIX)
        $(FC) $(FFLAGS)$(DEBUG)  $(INCS) -c makeparam$(SUFFIX)

makeparam$(SUFFIX): makeparam.F main.F
#
# MIND: I do not have a full dependency list for the include
# and MODULES: here are only the minimal basic dependencies
# if one strucuture is changed then touch_dep must be called
# with the corresponding name of the structure
#
base.o: base.inc base.F
mgrid.o: mgrid.inc mgrid.F
constant.o: constant.inc constant.F
lattice.o: lattice.inc lattice.F
setex.o: setexm.inc setex.F
pseudo.o: pseudo.inc pseudo.F
poscar.o: poscar.inc poscar.F
mkpoints.o: mkpoints.inc mkpoints.F
wave.o: wave.F
nonl.o: nonl.inc nonl.F
nonlr.o: nonlr.inc nonlr.F

$(OBJ_HIGH):
        $(CPP)
        $(FC) $(FFLAGS) $(OFLAG_HIGH) $(INCS) -c $*$(SUFFIX)
$(OBJ_NOOPT):
        $(CPP)
        $(FC) $(FFLAGS) $(INCS) -c $*$(SUFFIX)

fft3dlib_f77.o: fft3dlib_f77.F
        $(CPP)
        $(F77) $(FFLAGS_F77) -c $*$(SUFFIX)

.F.o:
        $(CPP)
        $(FC) $(FFLAGS) $(OFLAG) $(INCS) -c $*$(SUFFIX)
.F$(SUFFIX):
        $(CPP)
$(SUFFIX).o:
        $(FC) $(FFLAGS) $(OFLAG) $(INCS) -c $*$(SUFFIX)

# special rules
#-----------------------------------------------------------------------
# these special rules are cummulative (that is once failed
#   in one compiler version, stays in the list forever)
# -tpp5|6|7 P, PII-PIII, PIV
# -xW use SIMD (does not pay of on PII, since fft3d uses double prec)
# all other options do no affect the code performance since -O1 is used

fft3dlib.o : fft3dlib.F
        $(CPP)
        $(FC) -FR -lowercase -O2 -c $*$(SUFFIX)

fft3dfurth.o : fft3dfurth.F
        $(CPP)
        $(FC) -FR -lowercase -O1 -c $*$(SUFFIX)

fftw3d.o : fftw3d.F
        $(CPP)
        $(FC) -FR -lowercase -O1 -c $*$(SUFFIX)

wave_high.o : wave_high.F
        $(CPP)
        $(FC) -FR -lowercase -O1 -c $*$(SUFFIX)

radial.o : radial.F
        $(CPP)
        $(FC) -FR -lowercase -O1 -c $*$(SUFFIX)

symlib.o : symlib.F
        $(CPP)
        $(FC) -FR -lowercase -O1 -c $*$(SUFFIX)

symmetry.o : symmetry.F
        $CPP)
        $(FC) -FR -lowercase -O1 -c $*$(SUFFIX)

wave_mpi.o : wave_mpi.F
        $(CPP)
        $(FC) -FR -lowercase -O1 -c $*$(SUFFIX)

wave.o : wave.F
        $(CPP)
        $(FC) -FR -lowercase -O1 -c $*$(SUFFIX)

dynbr.o : dynbr.F
        $(CPP)
        $(FC) -FR -lowercase -O1 -c $*$(SUFFIX)

asa.o : asa.F
        $(CPP)
        $(FC) -FR -lowercase -O1 -c $*$(SUFFIX)

broyden.o : broyden.F
        $(CPP)
        $(FC) -FR -lowercase -O2 -c $*$(SUFFIX)

us.o : us.F
        $(CPP)
        $(FC) -FR -lowercase -O1 -c $*$(SUFFIX)

LDApU.o : LDApU.F
        $(CPP)
        $(FC) -FR -lowercase -O2 -c $*$(SUFFIX)

谢谢
回复此楼

» 猜你喜欢

» 本主题相关价值贴推荐，对您同样有帮助:
1楼 2012-04-03 23:11:22
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖

huangyc

木虫 (正式写手)
我不是想测试两个版本的速度区别，只是5.2编完了速度明显偏低。不知道vasp的计算速度与哪些编译参数有关系，比如DCACHE_SIZE的大小或中sacaLAPACK等
如何调整使得vasp运行时全部都是“R"
赞一下 回复此楼
3楼2012-04-04 15:06:06
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖
查看全部 3 个回答

souledge

专家顾问 (著名写手)

【答案】应助回帖

★ ★
感谢参与，应助指数 +1
fzx2008: 金币+2, 专家考核, 谢谢指教 2012-04-04 00:53:13
差不多配置不能作为标准，甚至偶尔一台机器的散热有问题也会影响速度～建议在同一台机器上使用同样的库和编译器编译VASP，然后进行对比～
我自己的测试结果，是4.6.28和5.2.2的速度基本没什么区别，尽管4.6.28使用的1级的优化，而5.2.2使用了3级优化～
赞一下 回复此楼
思想重于技巧，内涵重于表象
2楼2012-04-04 00:20:12
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖
查看全部 3 个回答
普通表情 龙 兔 虎 猫
最具人气热帖推荐 [查看全部] 作者 回/看 最后发表
[考研] 生物学学硕，一志愿湖南大学，初试成绩338 +5 YYYYYNNNNN 2026-03-26 6/300 2026-03-28 17:50 by 狂炫麦当当
[考研] 275求调剂 +10 Micky11223 2026-03-25 14/700 2026-03-28 15:48 by Micky11223
[考研] 329求调剂 +6 星野? 2026-03-26 6/300 2026-03-28 14:14 by 唐沐儿
[考研] 311求调剂 +4 冬十三 2026-03-24 4/200 2026-03-28 13:17 by 唐沐儿
[材料工程] 一志愿C9材料与化工专业总分300求调剂 +8 曼111 2026-03-24 9/450 2026-03-28 07:58 by YYYYX1234
[考研] 0856材料化工调剂 总分330 +10 zhubinhao 2026-03-27 10/500 2026-03-28 03:34 by fmesaito
[考研] 一志愿211，335分，0856，求调剂院校和导师 +4 倾____萧 2026-03-27 5/250 2026-03-27 11:52 by zhshch
[考研] 276求调剂。有半年电池和半年高分子实习经历 +10 材料学257求调剂 2026-03-23 11/550 2026-03-27 10:13 by YCIT- LHL
[考研] 081200-11408-276学硕求调剂 +4 崔wj 2026-03-26 4/200 2026-03-27 08:04 by chemisry
[考研] 284求调剂 +11 junqihahaha 2026-03-26 12/600 2026-03-27 04:37 by wxiongid
[考研] 304材料求调剂 +4 钟llll 2026-03-26 4/200 2026-03-27 03:42 by wxiongid
[考研] 333求调剂 +6 wfh030413@ 2026-03-23 6/300 2026-03-26 22:45 by 学员8dgXkO
[考研] 0703化学求调剂 +3 丹青奶盖 2026-03-26 5/250 2026-03-26 20:11 by macy2011
[考研] 086000生物与医药292求调剂 +6 小小陈小小 2026-03-22 9/450 2026-03-26 15:58 by dick_runner
[考研] 生物技术与工程 +3 1294608413 2026-03-25 4/200 2026-03-25 18:02 by 1294608413
[考研] 300分，材料，求调剂，英一数二 +5 超赞的 2026-03-24 5/250 2026-03-24 21:07 by 星空星月
[考研] 292求调剂 +4 鹅鹅鹅额额额额� 2026-03-24 4/200 2026-03-24 16:41 by peike
[基金申请] 请教下大家 2026年国家基金申请是双盲审吗？ +3 lishucheng1 2026-03-22 5/250 2026-03-24 08:22 by gltch
[考研] 384求调剂 +3 子系博 2026-03-22 6/300 2026-03-23 21:45 by 子系博
[考研] 269求调剂 +4 我想读研11 2026-03-23 4/200 2026-03-23 21:25 by pswait
信息提示
关闭
请填处理意见
关闭