🧠 Understanding CPU Usage for Mesh Generation and Analysis for MPIC

When working with MPIC (Multiphysics for IronCAD) simulation software, CPU resources play a critical role in determining performance. This article outlines how many CPU cores and threads are utilized during mesh generation and analysis across different software tiers: Free, Trial, Basic, and Advanced.

🧩 Mesh Generation

During the mesh generation phase, the number of CPU cores allocated depends on the license level:

• Core Count = Thread Count (if hyperthreading is off)
• Thread Count = Core Count × 2 (if hyperthreading is on)

Maximum core/thread usage per tier:

• Free: up to 2 cores
• Trial: up to 6 cores
• Basic: up to 2 cores
• Advanced: up to 6 cores

🧪 Analysis (Solve Phase)

The analysis stage (solver execution) is more dynamic in its resource usage. It uses an internal thread pool with a default maximum size of 640 threads. However, the actual number of cores/threads used depends on the system capabilities.

The solver uses a setting called Smax, which refers to the maximum core/thread count available on the user's machine. On most systems:

• If hyperthreading is off, threads = cores
• If hyperthreading is on, threads = cores × 2

🆕 Note on GPU Solver (2026 and beyond):
Starting in 2026, the solver supports GPU acceleration. In GPU-enabled environments, the computational workload is distributed across CUDA cores rather than CPU cores.
For example, NVIDIA's Blackwell architecture features approximately 68,000 CUDA cores, dramatically increasing solver throughput compared to CPU-only setups.

🧾 CPU Core & Thread Usage Table

*Actual thread count depends on whether hyperthreading is enabled on the system.