For large cases, such as cases that contain a large number of cell zones, you can utilize multiple CPU processes in parallel to reduce both the setup and postprocessing time for your simulaton. Multiple CPU processes for setup and postprocessing can be specified in the Fluent Launcher or through the command prompt, as outlined in Starting the Fluent GPU Solver.

After reading in a mesh or Fluent case file, you must partition or subdivide the mesh into groups of cells that will be solved on each CPU processor. For detailed information on partitioning the mesh see Mesh Partitioning and Load Balancing. Upon starting the calculation, the partitions created for setting up the case will be combined into a single partition for each GPU being used by the Fluent GPU Solver.