The Intel Distribution for LINPACK Benchmark supports command-line parameters for HPL that help you to avoid making small changes in the HPL.dat input file every time you do a new run.
Placeholders in this command line illustrate these parameters:
xhpl.exe -n <problem size> -m <memory size in Mbytes> -b <block size> -p <grid row dimn> -q <grid column dimn>
You can also use command-line parameters with the sample runme script. For example:
runme_intel64_dynamic.bat -m <memory size in Mbytes> -b <block size> -p <grid row dimn> -q <grid column dimn>
For more command-line parameters, see Heterogeneous Support in the Intel Distribution for LINPACK Benchmark.
If you want to run for N=10000 on a 1x3 grid, execute this command, provided that the other parameters in HPL.dat and the script are correct:
runme_intel64_dynamic.bat -n 10000 -p 1 -q 3
By using the m parameter you can scale by the memory size instead of the problem size. The m parameter only refers to the size of the matrix storage. Therefore, to use matrices that fit in 50000 Mbytes with NB=256 on 16 nodes, adjust the script to set the total number of MPI processes to 16 and execute this command:
runme_intel64_dynamic.bat -m 50000 -b 256 -p 4 -q 4