This document outlines the recommended hardware specifications for running Ansys Electronics Desktop, focusing on CPU, memory, storage, GPU, and other critical system components.
Optimizing Ansys Electronics Desktop Performance
Choosing the right hardware to use with Ansys tools will significantly impact productivity in terms of model size, performance, and user experience. This page is a resource for engineers looking to make informed hardware decisions specifically for Ansys Electronics Desktop (AEDT).CPU (Processor)
- Latest Generation CPUs with High Frequency: The performance of Ansys Electronics Desktop benefits significantly from high clock speeds. For instance, HFSS is an in-core solver where both clock speed and RAM speed are crucial[1][2][5]
- Number of Cores: The number of cores impacts how quickly a model can be solved. A good metric for comparison is (Clock Speed * Number of Cores / Cost). More cores can help with parallel processing and distributed solving [1][5][4].
Memory (RAM)
- Recommended RAM: It is recommended to have 8 GB per core, with typical systems having 64 GB to 256 GB[1][2][5].
- Memory Bandwidth: High-frequency RAM (e.g., 2933 MHz or higher) is important to ensure efficient data handling and performance. Ensure all memory channels in both processors are populated equally to operate at maximum speed [1][5].
Storage
- SSD (Solid State Drives): An SSD is highly recommended for all simulation work to improve read/write speeds and overall system responsiveness. NVMe SSDs are preferred when possible, with intermediate endurance levels (~0.3 DWPD) [1][4][5].
- Storage Configuration: Ansys recommends having a high-performance file system, preferably using SSDs with NVMe interface, configured in RAID0 for optimal performance. A combination of one smaller, faster drive (NVMe) for solving and one larger, slower drive (HDD) for storage is suggested [1][5][4].
Graphics and GPU Acceleration
Several Ansys Electronics Desktop solvers can benefit from GPU acceleration:
- HFSS frequency domain, transient, and SBR+ solvers.
- HFSS 3D Layout for matrix solves and eye analysis.
- Maxwell 3D eddy current matrix solves.
- Circuit Design for eye analysis.
- Circuit Netlist for eye analysis.
- EMIT
Electronics - GPU selection is based on the type of solve
SBR+ solvers can benefit from single precision GPU's with higher memory bandwidth, such as NVIDIA RTX series.
All other solvers require GPU's that are capable of double precision floating point calculations, such as NVIDIA A100 or 100 or P100 series.
If installing multiple GPU's make sure all GPUs are from the same series and as adequate level CUDA installed and GPU's set to computational mode
PCle4
CPU RAM at least 2x of the accumulated GPU RAM
Enough power from PSU
GPU Recommendations:
- For GPU acceleration, it is recommended to use NVIDIA Data Center GPUs of the Ampere series and Tesla GPUs of the Volta, Pascal, and Kepler generations. NVIDIA Workstation RTX and Quadro GPUs are also supported but with some exceptions [3][2][4].
- Discrete Graphics Cards: Use of discrete workstation graphics cards with at least 1 GB of discrete video memory supporting OpenGL version 4.5 or higher. For full functionality, a recent NVIDIA or AMD Professional or Workstation Graphics card is recommended [2][3][4].
Specific Requirements for Different Solvers:
- HFSS Frequency-domain and Time-domain solvers support NVIDIA Data Center GPUs and Tesla GPUs.
- HFSS SBR+ solver supports both Data Center GPUs and workstation GPUs.
- Maxwell solvers and Ansys EMIT also support a similar range of GPUs [2][3][4].
Other Considerations
- High-Speed Interconnects: For clustering or running on multiple nodes, high-speed interconnects like QDR or FDR IB are recommended. A minimum 10Gb network is required for efficient data transfer and communication between nodes[5][4][1]
- 3D Mouse: For navigating complex models, a 3D mouse can improve user interaction and efficiency [1]
- Multiple Monitors: Engineers tend to work with multiple monitors to efficiently move between applications and manage complex simulations[1]
Operating System and Drivers
- Supported Operating Systems: Ansys Electronics Desktop supports both Windows and Linux operating systems. Ensure the use of the latest drivers for the best performance. -NVIDIA drivers should be downloaded directly from the NVIDIA website to ensure compatibility and performance [2][3][4].
- Special Considerations for Quad-Core Sockets: Ansys generally does not recommend quad-core socket systems due to memory bandwidth issues. Consult Ansys support for guidance if considering such a configuration[1][2][3][4]
Conclusion
Choosing the right hardware for Ansys Electronics Desktop can significantly impact your productivity and simulation performance. Prioritize high-frequency CPUs, ample and fast RAM, SSD storage, and compatible discrete GPUs for the best experience.
For more detailed information, refer to the specific sections in the provided manuals and documents below.
References and Additional Resources
[1] Ansys System Hardware Requirements - Ozen Engineering, Inc
[2] Ansys 2024 R2 GPU Compute Capabilities.pdf
[3] Ansys 2024 R2 GPU Accelerator Capabilities.pdf
[4] Ansys 2024 R2 Graphical Display Graphics Cards Tested.pdf
[5] Ansys High-Performance Computing Tools.pdf
[6] 2041327-R2022.x-hardware-all.pdf
[7] 2041327-R2021.x-hardware-all.pdf
[8] Ansys Platform Support Strategy Plans January 2024.pdf