Changes between Version 7 and Version 8 of Public/User_Guide/Offloading_hybrid_apps

Timestamp:

Sep 17, 2019, 2:55:07 PM (5 years ago)

Author:

Kevin Sala

Comment:

—

Public/User_Guide/Offloading_hybrid_apps

@@ -157 +157 @@
 simulated). For this reason, the application provides `-g` option in order to
 control the maximum number of GPU processes. By default, the number of GPU processes
-will be half of the total number of processes.
+will be half of the total number of processes. Also note that the non-CUDA variants
+cannot compute kernels on the GPU. In these cases, the structure of the application
+is kept but the CUDA tasks are replaced by regular CPU tasks.
 
-Also note that the non-CUDA variants cannot compute kernels on the GPU. In this
-cases, the structure of the application is kept but the CUDA tasks are replaced
-by regular CPU tasks.
+Finally, the OpenMP variants can be executed similarly, but setting the `OMP_NUM_THREADS`
+to the corresponding number of CPUs per process. As an example, we could execute the
+following command:
 
+{{{#!bash
+$ OMP_NUM_THREADS=24 srun -n 4 -c 24 ./nbody.tampi.omp.2048bs.bin -t 100 -p 8912 -g 2
+}}}
 
 == References ==