Hardware Recommendations FAQ

How to Choose Hardware for your Simulation

  • The number of cores are generally decided based on the model size, model fidelity and expense.
    • Model size: Cells, grid points, nodes, elements, degrees of freedom
    • Model fidelity: Time accurate, Single vs Double precision and Solver
    • Cost focus: Minimize dollar cost per solution
    • Cost-performance focus: Good balance between time to solution and computing cost
  • For example, CFD generally start with 75K – 100K cells per core
  • A good rule of thumb is to initially utilize a smaller number of cores to test your model before running the full intended core count
UsageCore TypesProcessorClock SpeedMemory/nodeStorage/node
General PurposeEmerald
Ferrite
Luna
Amber		Intel Xeon Platinum P-8124
Intel Xeon Platinum 8168
Intel Xeon SP 2nd Gen
AMD EPYC 7551		3.0 GHz
2.7 GHz
3.6 GHz
2.0 GHz		144 GB
64 GB
192GB
240 GB		1.3 TB
256 GB
1.73 TB
700 GB
High MemoryZinc
Carbon
Melanite
Ruby		Intel Xeon E5-2686 v4
Intel Xeon Platinum 8168
Intel Xeon Platinum 8175
Intel Xeon Platinum 8175M		2.3 GHz
2.7 GHz
2.5 GHz
2.5 GHz		488 GB
352 GB
384 GB
768 GB		1.3 TB
700 GB
3.84 TB
3.6 TB

This information can be found on the Hardware Settings section for a job.

For instance, Emerald has 36 cores per node with 144.0 GB memory and 1.30 TB storage per node. Melanite, which is a High Memory Core Type, has 48 cores per node. In addition, Melanite has more memory, storage and cores per node compared to Emerald.

These details will help you determine the right core type for your simulation.

For CFD/FEA based simulations

  • Emerald
  • Amber

For FEA codes with higher memory requirements

  • Melanite
  • Zinc
  • Topaz (for higher memory requirements)

Other FAQs