a)Flops
b)Budget
c)Application.
Depends on the criteria we have to build the HPC Cluster Data center.
2) Performance should calculate like
a)Performance per $
b)Performance per speed
c)Performance per watt
3)NVIDIA GPU is used in the following criteria
1)Bioinformatics
2)Computational Finance
3)Computational Fluid Dynamics (CFD)
4)Electronic Design Automation.
5)Molecular Dynamics
6)Medical Imaging
7)Quantum Chemistry
8)Whether Climate
9)Match MATLAB.
4)HPC Important Application Name
1)Amber
2)Gromacs (molecular dynamic simulation)
3)Gaussian,
4)matlab,
5)OpenFoAM - Fluid Dynamic
6)NAMD,
7)LAMMPS,
AMBER 11 is the Biochemist -(Molecular Dynamics Research) HPC software. it is Bio molecular simulation package. HPC List of Software.
http://www.hpc.cam.ac.uk/user/software.html
Basic Requirement To Run job on Distributed Mode
1)NFS2)NIS 3)ssh passworless access (across the compute node)
2)When job is running if it is not progress after 1 minte , then there is issue with laughing distributing job - check all the node it is running the process - Check ssh on both client/server side.
3)if job is running around one hour/two hour then it is not progress then there is issue with solver(like internal programming error- check the out file.)
I got the following error
This is expected when FORTRAN error was encountered, otherwise
it can be caused by a very busy system or slow connection
to the network drive. OptiStruct will now exit.
Unless there is a specific error message in the out file
all computations and result files are already completed,
but some scratch files may require manual cleaning.
Signal 14 :: SIGALRM
In the out file i got like
*** INTERNAL PROGRAMMING ERROR ***
in file "newahashc.c", at location # 1040.
HSAS READ ERROR: 129195312
0 Comments