Ansys Electronics Desktop (AEDT) can significantly reduce simulation time by leveraging multiple CPU cores and GPU acceleration. However, enabling additional computational resources requires proper High-Performance Computing (HPC) licenses in addition to your base solver licenses.
By default, the base license allows the use of up to 4 cores without any additional HPC licensing. With Ansys HPC licenses, you can unlock more cores for your simulations. As shown below, the first HPC Pack enables 8 additional cores, the second increases the total to 36 cores, the third to 132 cores, and so on. This scalable approach provides a cost-effective way to access a high number of cores. For example, running a simulation on 2000 cores instead of 100 would require only two additional HPC Packs.
Performance Benefits of Multi-Core Processing
The performance benefits of multi-core processing are substantial, as shown in practical simulation results. In the example below, a complex SO-DIMM to UDIMM connector analysis with 2.3 million tetrahedra and a 10 million-size matrix achieved over a 5x speedup when scaling from a single core to 10 CPU cores. This significant improvement highlights the effectiveness of parallel processing in large electromagnetic simulations, where computational workloads can be efficiently distributed across multiple cores to greatly reduce solution time.
SO-DIMM to UDIMM connector courtesy of SMART Modular Technologies
In HFSS 3D Layout, high-performance computing delivers excellent scalability, significantly shortening simulation times for complex PCB designs.
A Maxwell 3D simulation further demonstrates this scalability, achieving a 70x speedup and reducing simulation time from 15 days to just 5 hours.
GPU Acceleration in AEDT
In addition to CPU scaling, GPU acceleration provides a “turbo boost” for finite element simulations in AEDT. It is particularly effective in handling the largest frontal matrices, delivering significant performance improvements, as shown below.
GPUs provide substantial acceleration for two key solver technologies in HFSS and Maxwell, each optimized for different computational challenges.
In HFSS Transient, the GPU-accelerated hybrid solver significantly outperforms traditional implicit solvers, particularly for high-frequency problems with uniform meshes and high operating frequencies. It typically delivers strong speedups starting around 140K degrees of freedom. If a project is not suitable for GPU acceleration, such as when the expected speedup is less than 1x, the solver automatically defaults to CPU processing.
For frequency-domain simulations, GPUs also accelerate the multifrontal sparse direct solver in both HFSS and Maxwell. This solver achieves meaningful performance improvements starting at around 2 million degrees of freedom. As with the hybrid solver, GPU resources are used only when a performance benefit is expected, ensuring efficient use of computational resources across different problem sizes and complexities.
Performance benchmarks show the considerable benefits of GPU acceleration across a range of electromagnetic simulation tasks.
Another Maxwell 3D example shows that larger matrix sizes yield significant speedups, further highlighting the value of GPU-enabled workflows for large-scale problems.
Enabling Additional Cores
To enable additional CPU cores in AEDT:
1. Go to the Simulation tab and click on HPC Options. This opens the HPC and Analysis window.
2. Select the design type for which you want to change the number of cores.
3. Click Edit to open the Analysis Configuration window.
4. Change the number of cores, then click OK to apply the changes.
HFSS
Maxwell 3D Maxwell 2D and Q3D
For other solvers such as AEDT Icepak, and EMIT, the steps to enable additional cores follow the same procedure described above.
GPU Acceleration
Several solvers in AEDT support GPU acceleration to improve simulation performance:
- HFSS frequency domain, transient, and SBR+ solvers
- HFSS 3D Layout for matrix solves
- Maxwell 3D for eddy current matrix solves
- EMIT
Each GPU typically requires 8 HPC units. The type of HPC license selected determines which license is used and how HPC units are consumed.
For solvers that support GPU acceleration, you can configure GPU settings in the Analysis Configuration window.
To enable GPU acceleration:
1. Go to Tools > Options > HPC and Analysis Options, and select the Options tab. Choose the design type for which you want to configure GPU settings.
2. Set Enable GPU to True by clicking the value next to it and toggling it on. Note that HFSS includes a separate toggle for Enable GPU for SBR+ Solve.
3. Go to the Configurations tab and click Edit.
4. In the Analysis Configuration window, uncheck "Use Automatic Settings" to access manual options and specify the number of GPUs to be used.
Note that GPU acceleration is only utilized when the model exceeds 2 million tetrahedra. For smaller models, AEDT will not use the GPU.
By following these steps, you can effectively leverage additional CPU cores and GPU acceleration across all Ansys Electronics Desktop tools to significantly reduce simulation time.
The video below shows how to enable additional CPU cores and GPU acceleration in AEDT, as outlined above.
Ozen Engineering Expertise
Ozen Engineering Inc. leverages its extensive consulting expertise in CFD, FEA, optics, photonics, and electromagnetic simulations to achieve exceptional results across various engineering projects, addressing complex challenges such as antenna design, signal integrity, electromagnetic interference (EMI), and electric motor analysis using Ansys software.
We offer support, mentoring, and consulting services to enhance the performance and reliability of your electronics systems. Trust our proven track record to accelerate projects, optimize performance, and deliver high-quality, cost-effective results. For more information, please visit https://ozeninc.com.
If you want to learn more about our consulting services, please visit: https://www.ozeninc.com/consulting/
CFD: https://www.ozeninc.com/consulting/cfd-consulting/
FEA: https://www.ozeninc.com/consulting/fea-consulting/
Optics: https://www.ozeninc.com/consulting/optics-photonics/
Photonics: https://www.ozeninc.com/consulting/optics-photonics/
Electromagnetic Simulations: https://www.ozeninc.com/consulting/electromagnetic-consulting/
Thermal Analysis & Electronics Cooling: https://www.ozeninc.com/consulting/thermal-engineering-electronics-cooling/