Centre of Excellence in Bioinformatics(CoE)
Tesla Usage Chart
Tesla Job Registration
Instructions for running simulations under Tesla Server
Gromacs Version 5.1
(Please check the manual for running jobs)
| 1. Before running simulations, the structure should be a properly validated one (Whatif) and report for the same has to be submitted along with your registration under http://tesla.bicpu.edu.in. |
| 2. Job commands must be executed by prefixing “nohup” and by suffixing “&” commands Ex. GROMACS 2019 (FALCON) & 5.1.2 (TESLA) nohup gmx mdrun -s xxx.tpr -cpi xxx.cpt -v -nb gpu & |
| GROMACS 5.1.2 (FALCON) nohup mpirun -np 10 gmx_mpi mdrun -ntomp 4 -cpi state_prev.cpt -noappend -v -deffnm md_100ns & |
| 3. Users are allowed to run jobs only for 10 days (with the dual Tesla GPU C2050 cards user can run 50ns to 100ns jobs with respect to number of residues and molecules added for simulations) |
| 4. After completion of jobs user data will be kept in the server for a maximum period of 7 days after which data will be deleted automatically. So it is advised to take back up once the user’s job get over in the server. |
| 5. User can run any number of jobs within the allocated time. The suffixed GPU command with Gromacs command will automatically perform better paralyzed calculation with the dual GPU cards. |
| 6. Users in the queue are requested to prepare input data files compatible with Gromacs 5.1 version to avoid time delay. And it is also advised to complete all the preliminary steps like pdb to gromacs conversion, energy minimization and solvation etc., in the user’s local system itself so the final mdrun step can be performed in the Tesla server. |
| 7. If any Power / Hardware failure or any other technical issues, the down time will be estimated and it will be added to the user to extend their job. |
Note: All the users who are running job under the Tesla Server, should acknowledge the server where the result being published.