Policies

In order to efficiently use our computational resources, we ask all group members to follow the guidelines below when planning and running simulations:

• Please run jobs on the computational nodes ("slave nodes"), rather than the head node, of each cluster. In the past, head node crashes – a great inconvenience to everyone – have occurred when all eight of its processors are engaged with computationally intensive work.
• Please do not run jobs "on top of" those of other users. If a node is fully occupied and you need to run something, please contact the other user, rather than simply starting your job there.
• For fastest disk read/write speeds, write the to local /scratch/username/ directory on each node, rather than your home directory. Home directories, which are accessible from every node, are physically located on the head node, so that reading and writing to disk may be limited by network transmission rates.
• The new cluster, with its fast interconnection hardware, is very well suited for large-scale simulations which benefit from the use many processors. A queuing system will be used to manage jobs on this cluster, and no jobs should run outside this queue system.

Please note that we attempted to implement the OpenPBS queue system on Cyrus1 and Quantum2 in December 2009; these systems appeared to be working in testing, but did not perform as desired when multiple jobs were submitted. The use of these queuing systems on those clusters has been suspended until further notice.

Computer Clusters

Our three clusters are Darius, Cyrus1, and Quantum2. See below for additional information. For information about using Gaussian at MIT, see below.

Darius

Darius is our newest cluster, with 30 computational nodes, each with 8 virtual processors.

More details will be posted soon.

Cyrus1

Hostname: cyrus1.csbi.mit.edu
Cyrus1 is a 24-node cluster, installed December 2008.

Please note that the CSBi network, on which Cyrus1 is hosted, does not allow access from IP addresses external to MIT. For remote access to Cyrus1, see the MIT IST VPN site.

Status notes

Node 10 (n010) is currently inaccessible.

Node information

The table below lists the memory and swap file size of each node on Cyrus1, along with the amount of space used and free in each node's local /scratch/ directory. This information is current as of April 1, 2010.

Quantum2

Hostname: quantum2.mit.edu
Quantum2 is a 20-node cluster installed in October 2007, which features high-memory nodes for quantum-chemical calculations.

Available Software

VMD 1.8.7 has been installed, and is working with a user's local Xwindows server. The executable is "/usr/local/vmd-1.8.7" (actually not a binary executable, but a script), and can be invoked using "vmd-1.8.7."

A number of software packages have been installed in /home/gpw501/software/
Please contact gwood@mit.edu for any questions; a very limited usage guide is given below.

GAMESS-US quantum chemistry package.

    Usage:  /home/gpw501/software/gamess/rungms JOB VERNO NCPUS >& JOB.log &
            JOB is the name of JOB.inp file to be executed
            VERNO is the current version of gamess (01 at the time of writing)
            NCPUS number of cpus   
      you must make a scratch directory that matches your user name on the
      node you are running from i.e. /scratch/$USER

See GAMESS homepage for more details.

CPMD (QMMM version) plane-wave/PP quantum CP and BO dynamics.

    Usage:  mpirun -n NCPUS cpmd.x JOB PATH-TO-PPs >& JOB.out &
            mpirun should be set to /opt/openmpi/tcp-gnu/bin/mpirun in your .bashrc
            NCPUS number of cpus
            JOB name of job file to be executed 
            PATH-TO-PPs the path to a pseudo potential library (see for eg /home/gpw501/software/cpmd/pseudos/ )
            JOB.out name of output file

Amber 10 molecular dynamics program. Executables are located in /home/gpw501/software/amber10/exe/

Gromacs molecular dynamics program. Was installed as root so binaries for MD and the gromacs tools are located in /usr/local/bin/.

propka electrostatic and pka computations for proteins. See /home/gpw501/software/propka2.0src/README_PROPKA2.0

Status notes

Nodes 4 and 14 (n004 and n014) are currently inaccessible, due to apparent hard disk problems.

Node information

The table below lists the memory and swap file size of each node on Quantum2, along with the amount of space used and free in each node's local /scratch/ directory. This information is current as of April 1, 2010.

Gaussian03 at MIT

MIT currently has a site license to run Gaussian03 on IS&T-owned computers, such as those in Athena clusters. For more information, see Gaussian on Athena. For information about creating/submitting Gaussian jobs, see these notes from the 10.675J course taught at MIT: Basics of Running G03.