03_shared-memory/01_bandwidth

Hands-on: Measure memory bandwidth

• Choose either C/C++ (c and ctrad) or Fortran (fortran). Both of them are fine, as well.

C/C++

How to execute

1. Edit a job script

• Before trying this hands-on, you need to do 00_stream to compile STREAM.
• We have three working directories, c/fj_zfill, c/fj, and ctrad/fj_zfill. Under each of directories, you can find:
  • run.sh : a script to execute STREAM
  • task.sh: a job script to run STREAM with different kinds of settings
• edit task.sh (modify --gname option).
• Edit BINDIR variable in run.sh before the execution. You need to write your installed location of STREAM binary (e.g., stream.exe) there.

2. Run program

• You can run the program either:

## Here is an example of c/fj_zfill.
## To run as a batch job
$ cd c/fj_zfill
$ pjsub task.sh 
## Or, to run in an interactive job
$ cd c/fj_zfill
$ bash task.sh 

• Each of the cases in the Exercises will be completed within 3-4 minutes.
  • For safety, we set the job elapsed time in the job scripts is 6 minutes.

Exercises A

• E1: Check task.sh in c/fj_zfill, for example. You can find that the number of OpenMP threads varies from 1 to 48 via NTHREADS variable.
• E2: Run the STREAM benchmark (stream.exe) in c/fj_zfill/,c/fj/, and ctrad/fj_zfill. Check the results in Best Rate (MB/s) of functions Add and Triad in 48 threads.
• E3: Compare the measurement results to the peak performance of memory bandwidth in A64FX.

Fortran

How to execute

1. Edit a job script

• Before trying this hands-on, you need to do 00_stream to compile STREAM.
• We have two working directories, fortran/fj_zfill and fortran/fj. Under each of directories, you can find:
  • run.sh : a script to execute STREAM
  • task.sh: a job script to run STREAM with different kinds of settings
• edit task.sh (modify --gname option).
• Edit BINDIR variable in run.sh before the execution. You need to write your installed location of STREAM binary (e.g., stream.exe) there.

2. Run program

• You can run the program either:

## Here is an example of fortran/fj_zfill.
## To run as a batch job
$ cd fortran/fj_zfill
$ pjsub task.sh 
## Or, to run in an interactive job
$ cd fortran/fj_zfill
$ bash task.sh 

• Each of the cases in the Exercises will be completed within 3-4 minutes.
  • For safety, we set the job elapsed time in the job scripts is 6 minutes.

Exercises A

• E1: Check task.sh in fortran/fj_zfill, for example. You can find that the number of OpenMP threads varies from 1 to 48 via NTHREADS variable.
• E2: Run the STREAM benchmark (stream.exe) in fortran/fj_zfill/ and fortran/fj/. Check the results in Rate (MB/s) of functions Add and Triad in 48 threads.
• E3: Compare the measurement results to the peak performance of memory bandwidth in A64FX.