计算化学公社
标题:
求问,mpi并行计算周期性石墨烯的体系分解问题
[打印本页]
作者Author:
SDown
时间:
2025-5-29 17:01
标题:
求问,mpi并行计算周期性石墨烯的体系分解问题
版本
:2018.3
求问,系综NVT跑周期性石墨烯(盒子尺寸:5.15755 * 5.10480 * 5.20000 nm),出现了域分解(Domain Decomposition)问题。
报错:
Initializing Domain Decomposition on 32 ranks
Dynamic load balancing: locked
Minimum cell size due to atom displacement: 0.433 nm
Initial maximum inter charge-group distances:
two-body bonded interactions: 7.084 nm, Exclusion, atoms 927 1007
multi-body bonded interactions: 7.084 nm, Angle, atoms 927 1007
Minimum cell size due to bonded interactions: 7.793 nm
Guess for relative PME load: 0.23
Will use 24 particle-particle and 8 PME only ranks
This is a guess, check the performance at the end of the log file
Using 8 separate PME ranks, as guessed by mdrun
Scaling the initial minimum size with 1/0.8 (option -dds) = 1.25
Optimizing the DD grid for 24 cells with a minimum initial size of 9.741 nm
The maximum allowed number of cells is: X 0 Y 0 Z 0
复制代码
原子 927 和 1007 的
two-body bonded interactions
距离计算似乎有问题(7.084 nm比盒子边长还大)。
927 和 1007 的位置如图所示在盒子两端,它们之间不应该有键,top文件里这两个原子间也没有键,按照石墨烯的结构它们之间也不应该存在键。
(, 下载次数 Times of downloads: 24)
上传 Uploaded
点击下载Click to download
考虑到 7.084nm ≈ 5nm * sqrt(2),怀疑是把两个原子在盒子里的距离(7.084nm)而不是跨过周期性边界的距离(~0.2nm)当成了原子间距离。
在 mpd 文件中设置了周期边界条件,但不起作用,设置freeze也没有帮助
单机不会报分解错误,但freeze指令无法生效,石墨烯会平移;单节点单线程可以运行,但生成的结果文件无法读取
请问应该怎么解决?
相关mdp和sbatch指令如下:
title = OPLS NVT equilibration
define = -DPOSRES ; position restrain the protein
; Run parameters
integrator = md ; leap-frog integrator
nsteps = 40000000 ; 1 fs * 40000000 = 40 ns
dt = 0.001 ; 1 fs
; Output control
nstxout = 2000 ; save coordinates every 20 ps
nstvout = 2000 ; save velocities everys 20 ps
nstenergy = 2000 ; save energies every 20 ps
nstlog = 2000 ; update log file every 20 ps
; Bond parameters
continuation = yes ; first dynamics run
constraint_algorithm = lincs ; holonomic constraints
constraints = h-bonds ; bonds involving H are constrained
lincs_iter = 1 ; accuracy of LINCS
lincs_order = 4 ; also related to accuracy
; Nonbonded settings
cutoff-scheme = Verlet ; Buffered neighbor searching
ns_type = grid ; search neighboring grid cells
nstlist = 10 ; 20 fs, largely irrelevant with Verlet
rcoulomb = 1.0 ; short-range electrostatic cutoff (in nm)
rvdw = 1.0 ; short-range van der Waals cutoff (in nm)
DispCorr = EnerPres ; account for cut-off vdW scheme
; Electrostatics
coulombtype = PME ; Particle Mesh Ewald for long-range electrostatics
pme_order = 4 ; cubic interpolation
fourierspacing = 0.16 ; grid spacing for FFT
; Temperature coupling is on
tcoupl = Berendsen ; modified Berendsen thermostat
tc-grps = K SOL CL GRA ; two coupling groups - more accurate
tau_t = 0.1 0.1 0.1 0.1 ; time constant, in ps
ref_t = 300 300 300 300 ; reference temperature, one for each group, in K
; Pressure coupling is off
pcoupl = no ; no pressure coupling in NVT
; Periodic boundary conditions
pbc = xyz ; 3-D PBC
; Velocity generation
gen_vel = no ; assign velocities from Maxwell distribution
freezegrps =GRA
freezedim =Y Y Y
复制代码
#SBATCH --nodes=1
#SBATCH --ntasks-per-node=32
#SBATCH --get-user-env
source ~/.bashrc
srun hostname -s | sort -n > slurm.hosts
gmx grompp -f NVT.mdp -c GRA.gro -p GRA.top -o NVT.tpr
mpirun -n 32 -machinefile slurm.hosts gmx mdrun -v -s NVT.tpr -o NVT.trr -c GRA.gro
复制代码
欢迎光临 计算化学公社 (http://ccc.keinsci.com/)
Powered by Discuz! X3.3