BlogJava-大漠驼铃-随笔分类-Chemistry

Centos 安装 Cuda3.0

草原上的骆驼 — Tue, 23 Mar 2010 14:38:00 GMT

1、下载

下载显卡驱动
http://www.nvidia.cn/Download/index.aspx?lang=cn
选择与显卡一至的驱动

下载cuda toolkit
http://developer.nvidia.com/object/cuda_3_0_downloads.html

这里选择的是CUDA Toolkit for RedHat Enterprise Linux 5.3 64位的，因为系统为64位

同样在这个页面下载测试用的CUDA SDK，测试安装环境是否成功。
GPU Computing SDK code samples and more，同样是选择64位。

2 安装
运行 /sbin/init 3
进行命令行模式，如果一开始没有安装图形界面不需要运行此命令。
ubuntu的进行命令行模式命令为 /etc/init.d/gdm stop/start
或者用键盘 ctrl+alt+f1

安装Nvidia驱动
ssh NVIDIA-Linux-x86_64-195.36.15-pkg2.run
一路默认安装即可。

安装Cuda toolkit
ssh cudatoolkit_3.0_linux_64_rhel5.3.run

默认安装，安装完成后配置环境变量
vim /etc/profile (root 用户)
export PATH=$PATH:/usr/local/cuda/bin
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda/lib64:/usr/local/cuda/lib

退出VIM, source /etc/profile

进行普通用户，同样让/etc/profile生效(source /etc/profile)
安装 cudasdk
sh gpucomputingsdk_3.0_linux.run
一路默认即可，进行用户目录下的
cd NVIDIA_GPU_Computing_SDK/C
目录

make 命令

此时也许会报错
/usr/bin/ld: cannot find -lglut

yum install *glu*
会发现有 freeglut-dev包

安装
yum install freeglut-dev

然后再make即可成功。

草原上的骆驼 2010-03-23 22:38 发表评论

amber9单机版安装

草原上的骆驼 — Wed, 06 Jan 2010 08:10:00 GMT

1．安装

(1)把AMBER9.TGZ文件拷贝到/usr/local/

#cp–rf /media/AMBER9/AMBER.TGZ /usr/local/

(2)把AMBER9解压到/usr/local/amber9

#tar–zxvf AMBER9.TGZ

2．bugfix.all补丁 http://amber.scripps.edu下载

(1)bugfix.all拷贝到/usr/local/amber9目录下

#cp–rf bugfix.all /usr/local/amber9

(2)#cd /usr/local/amber9

#patch-p0 -N -r patch-rejects < bugfix.all

3．安装INTEL编译器intel fortran和intel c++

http://www.intel.com/cd/software/products/asmo-na/eng/282048.htm

http://www.intel.com/cd/software ... ers/clin/277618.htm

下载 l_cc_p_10.0.023_ia32.tar.gz 与 l_fc_p_10.0.023_ia32.tar.gz

解压文件并执行

#./install.sh ；按说明一步一步安装，默认安装在/opt/intel

安装libstdc++.so.5

#yum -y install compat-libstdc++-33

4．设置编译器环境

#vi/etc/profile

AMBERHOME=/usr/local/amber9

exportAMBERHOME

source/opt/intel/fc/10.0.023/bin/ifortvars.sh

source/opt/intel/cc/10.0.023/bin/iccvars.sh

保存并重启系统

5．进入AMBERHOME/src

#cd $AMBERHOME/src

#./configureifort_ia32

#makeserial

6．测试基本的AMBER

#cd$AMBERHOME/test

#maketest.serial

草原上的骆驼 2010-01-06 16:10 发表评论

pbs 书写

草原上的骆驼 — Tue, 08 Dec 2009 01:15:00 GMT

#!/bin/sh
#PBS -l nodes=1:ppn=8
#PBS -q students
#PBS -j oe
#PBS -N heme
cd /home/sxtian/cf3/CF3-Cl
/usr/local/chemsoft/autodock4.2.2.1/bin/autodock4 -p CF3-Cl.dpf -l CF3-Cl.dlg

test1@master t]vi pbsjob
#!/bin/tcsh
#PBS -o /home/test1/pbstest/t/output 标准输出文件
#PBS -e /home/test1/pbstest/t/error 错误输出文件
#PBS -l nodes=5:ppn=4 规定使用的节点数以及每个节点能跑多少核
#PBS –q students 把任务提交到students队列中
cd $PBS_O_WORKDIR 到工作目录下（此为PBS提供的环境变量）
mpirun –machine $PBS_NODEFILE -np 20 ./vasp

草原上的骆驼 2009-12-08 09:15 发表评论

PBS

草原上的骆驼 — Tue, 08 Dec 2009 01:14:00 GMT

pbs(1B)                               PBS                              pbs(1B)

NAME
       pbs - general information on pbs

DESCRIPTION
       PBS stands for Portable Batch System. It is a networked subsystem for
       submitting, monitoring, and controlling a work load of batch jobs on
       one or more systems. More information about PBS is available in the
       PBS Users Guide.

       Batch means that the job will be scheduled for execution at a time cho‐
       sen by the subsystem according to a defined policy and the availability
       of resources. For a normal batch job, the standard output and standard
       error of the job will be returned to files available to the user when
       the job is complete.   This differs from an interactive session where
       commands are executed when entered via the terminal and output is
       returned directly to the terminal.   PBS also supports an interactive
       batch mode where the input and output is connected to the user’s termi‐
       nal, but the scheduling of the job is still under control of the batch
       system.

       A job is typically a shell script and a set of attributes which provide
       resource and control information about the job. A job does not have to
       be submitted on the system where it will run, tt can be submitted on
       any system with the PBS commands and access to the execution system,
       see qsub(1B). Output will be returned to the system from which the job
       was submitted unless directed otherwise.

       Attributes offer control over when a job is eligible to be run, what
       happens to the output when it is completed and how the user is notified
       when it completes. The attributes of the job may be specified on the
       command line or in the job script when the job is submitted. For
       information    about    job    attributes,     see     qsub(1B)     and
       pbs_job_attributes(7B).

       One important attribute is the resource list. The list specifies the
       amount and type of resources needed by the job in order to execute.
       The list also implies a hard upper limit on usage of those resources.
       When the limit is reached, the job is terminated.   The types of
       resources available to a job vary with the system architecture. For a
       list   of   resources   supported   on   the   default   system,    see
       pbs_resources(7B).   There are man pages for other systems types as
       well,    see     pbs_resources_aix4(7B),     pbs_resources_fujitsu(7B),
       pbs_resources_irix5(7B),                    pbs_resources_solaris5(7B),
       pbs_resources_sp2(7B), pbs_resources_sunos4(7B), or pbs_resources_uni‐
       cos8(7B).

       Once a job has been submitted, it may be monitored by use of the
       qstat(1B) command. Two forms of output are available with the qstat
       command.   The default form is the short display.   Information about a
       job is limited to a single line.   Complete information about the job
       or jobs is available through qstat with the -f option. Information
       will be given about all jobs in the system, all jobs in specified
       queues, or only specified jobs.

       When displaying status of jobs, you will see in which queue the job
       resides. In PBS a queue is just a collection point for jobs, it does
       not imply any execution ordering. That ordering is determined by a
       scheduling policy implemented by the system administration.

       Other commands of interest which have man pages of their own are:

       qalter Alter a job’s attributes.

       qdel    Delete a job.

       qhold   Place a hold on a job to keep it from being scheduled for run‐
               ning.

       qmove   Move a job to a different queue or server.

       qmsg    Append a message to the output of an executing job.

       qrerun Terminate an executing job and return it to a queue.

       qrls    Remove a hold from a job.

       qselect Obtain a list of jobs that met certain criteria.

       qsig    Send a signal to an executing job.

SEE ALSO
       qalter(1B),   qdel(1B), qhold(1B), qmove(1B), qmsg(1B), qrerun(1B),
       qrls(1B), qselect(1B), qsig(1B), qsub(1B) and the PBS User Guide.
       Starting with qsub(1B), you can find all other available PBS man pages
       by following references in the "See Also" section.

Local                                                                  pbs(1B)

草原上的骆驼 2009-12-08 09:14 发表评论

qsub

草原上的骆驼 — Tue, 08 Dec 2009 01:14:00 GMT

摘要: qsub(1B) PBS &n... 阅读全文

草原上的骆驼 2009-12-08 09:14 发表评论