RAM and Core processors

kpk

Do we have any options to control System RAM and Core processor usage for Solving or Run in Acusolve/HW CFD?

Alexandru-Ionel Constantin

Hi KPK,

 

We can control the number of cores used by setting the Parallel processing from the Launch AcuSolve window to one of the three options:

- Serial -- use only one core, in the log file we see: Message passing type = none, Number of subdomains = 1 and Number of processors(cores) = 1;

- Intel MPI (local) - where we can set the numbers of processors(cores used)during the solving process) AcuSolve will decompose our domain(domain decomposition), in the log file we see -- Message passing type = OpenMP, Number of subdomains = x and Number of processors(cores) = x, where x is the number we set.

- Platform MPI(probably for a cluster/network)- where we can set the numbers of processors(cores used)during the solving process) AcuSolve will decompose our domain, in the log file we see -- Message passing type = OpenMP, Number of subdomains = x and Number of processors(cores) = x, where x is the number we set.

 

Domain decomposition is used to distribute the elements and nodes to different processors. Message Passing Interface (MPI) is used to communicate between distributed memory computers and shared memory data copy is used between subdomains of a single shared memory box. This ensures minimum communication costs. All I/O during the computation is done in parallel. AcuSolve performs profiling of each run to determine where the majority of the CPU/elapsed time is spent during the simulation. The results of this profiling are printed to the .log file at the conclusion of each run and can be reviewed to assess the performance of the code on your computing system.

 

We have no option to control the maximum RAM used.

 

Best regards,

Alexandru

kpk

Alexandru-Ionel Constantin said:

Hi KPK,

 

We can control the number of cores used by setting the Parallel processing from the Launch AcuSolve window to one of the three options:

- Serial -- use only one core, in the log file we see: Message passing type = none, Number of subdomains = 1 and Number of processors(cores) = 1;

- Intel MPI (local) - where we can set the numbers of processors(cores used)during the solving process) AcuSolve will decompose our domain(domain decomposition), in the log file we see -- Message passing type = OpenMP, Number of subdomains = x and Number of processors(cores) = x, where x is the number we set.

- Platform MPI(probably for a cluster/network)- where we can set the numbers of processors(cores used)during the solving process) AcuSolve will decompose our domain, in the log file we see -- Message passing type = OpenMP, Number of subdomains = x and Number of processors(cores) = x, where x is the number we set.

 

Domain decomposition is used to distribute the elements and nodes to different processors. Message Passing Interface (MPI) is used to communicate between distributed memory computers and shared memory data copy is used between subdomains of a single shared memory box. This ensures minimum communication costs. All I/O during the computation is done in parallel. AcuSolve performs profiling of each run to determine where the majority of the CPU/elapsed time is spent during the simulation. The results of this profiling are printed to the .log file at the conclusion of each run and can be reviewed to assess the performance of the code on your computing system.

 

We have no option to control the maximum RAM used.

 

Best regards,

Alexandru

Thanks for the useful information

Alexandru-Ionel Constantin

kpk said:

Thanks for the useful information

I’m glad I could help.

ydigit

Generally RAM would not be in issue for most users with AcuSolve. AcuSolve is not that RAM hungry  

It would be approx 1 million nodes (~6 million tetras) would require around ~3GB. 

Assuming you have a decent machine for CAE with 16GB RAM, it would be possible to run a case with ~5 Million Nodes (~30 Million Elements). So the main issue that you might face is the processor speed and number of cores available for running such a simulation. 

 

kpk

ydigit_21275 said:

Generally RAM would not be in issue for most users with AcuSolve. AcuSolve is not that RAM hungry  

It would be approx 1 million nodes (~6 million tetras) would require around ~3GB. 

Assuming you have a decent machine for CAE with 16GB RAM, it would be possible to run a case with ~5 Million Nodes (~30 Million Elements). So the main issue that you might face is the processor speed and number of cores available for running such a simulation. 

 

Thanks ydigit for the useful information