Learning Forum

[Chris Roberts]

Hello all,

I am running a simulation with around 5.5-6 million elements. It is a single phase ( fluid water ) laminar model. I was originally running this model on a 32 core threadripper, 256 GB Ram,  1TB Nvme Drive, ASUS wrx8 Motherboard. On this machine it took 24 hours to complete about 2500 iterations at least. We recently upgraded by adding a second PC with the same specs expect with a more power 64 core threadripper. All other specs are the same. However we noticed for the exact same simulation in 24 hours only 800-900 iterations complete and it runs extremely slow compared to the 32 core for the exact same simulation. Both PCs are using Ansys HPC academic licensing. If there are any suggestions as to why the more powerful CPU is running slower I would definitely appreciate it. Thank you for the help all.

[Nilay Pedram]

Hello,

Only increasing the cores alone won't help. If the core count is increased too much, it will take significant time to pass messages across the cores. and it may end up taking more time. Also, if the mesh is same then at one point you won't see a huge difference in the time even if you use multiple nodes. Apart from these, there can be another reason too for the slow down. 

For more information, please refer to Fluent User's Guide - Part III: Solution Mode (ansys.com)

If you are not able to access the above link, please follow this forum discussion - How to access the ANSYS Online Help

Hope this helps you. Thank you!