Please see the document called Platform Support by Application / Product on Ansys's Platform Support page for a full list of supported OS's for current and previous versions of Ansys.
 

General Recommendations

• Windows 10 - is the most commonly used and supported operating system.  This is the OS that SimuTech engineers and the majority of our customers use to run Ansys and what we recommend. 
• Windows 11 - As of Ansys version 2022 R2 and after, Windows 11 is supported. While Ansys versions supported on Windows 10 will likely work on Windows 11, we have not widely deployed or tested prior versions of Ansys on Windows 11 at this time and still officially recommend using Windows 10 for the time being if you need to use versions prior to 2022 R2.
• Linux - While not for the typical user, Ansys does also provide support for different Linux versions.  See the Platform Support page for more info.  
   

Unsupported Operating Systems

• Windows 7 (unsupported by Microsoft) & Windows 8 (mainstream support has ended) may run Ansys, however, they are officially unsupported and we have experienced software errors while running on these OS's. We strongly recommend upgrading to Windows 10 for compatibility, performance and security reasons.

• Mac OS is not supported and Ansys cannot be installed on this OS. While you should be able run Ansys on a Windows Virtual Machine inside Mac OS with a program like Parallels, this is NOT supported and your mileage may vary in regards to compatibility, and performance will be hindered.

Visit Ansys' Platform Support webpage for further details regarding supported OS's and 3rd party software & hardware that integrates with Ansys, such as:

• Summary of Future Strategy and Plans
• Supported Operating Systems by Ansys Application
• Supported CAD Software & Versions
  • Examples:  ACIS, AutoCAD, CATIA, Creo, Fusion 360, IGES, Inventor, JT, Monte Carlo N-Particle, NX, Parasolid, Revit, Rhinoceros, SketchUp, Solid Edge, Solidworks, SpaceClaim, STEP
• Supported Browsers
• Supported 3Dconnexion Devices
• Supported GPU Accelerator and Co-Processor Capabilities
• Supported Graphics Cards
  • Examples:
    • Nvidia - Quadro GV, Quadro P, Quadro RTX, Quadro T, RTX, T
    • AMD - Radeon Pro, Ryzen, Ryzen Pro
• Supported Interconnects
• Supported Job Schedulers & Queuing Systems Support
• Supported Remote Display Applications

CPU (Processor):

• Solver is aided by AVX512 support, thus Intel CPUs based on Skylake-SP and Skylake-X or newer are recommended. Check Intel ARK Database for AVX512 Support (but not Xeon Phi series).
• Examples of recommended CPUs for HPC applications: i9-9920X or Xeon Gold 6136 (12 cores), Xeon Gold 6140 (18 cores * 2 = 36 cores)
• High core frequency is ideal, but actual frequency is dependent or core use, instruction set use, and adequate cooling. WikiChip is an excellent resource on turbo speeds.

Memory (RAM):

• Memory capacity is very important. The model solution should be solving in RAM (“in-core”) and not paging to the hard drive (“out-of-core”). About 15 GB per million degrees of freedom is required, but this is approximate (45 GB per million solid nodes in structural analysis).
• Fastest memory (i.e. 2666 MHz for servers or 3200+ for i9 based desktops).
• To operate at maximum speed, all memory channels in both processors should be populated with equal amounts of memory.
• Note: Purchasing more RAM will alleviate I/O issues, as (1) the solver can use this memory to avoid doing as much I/O, and (2) the operating system can then use available RAM to cache or buffer these I/O that MAPDL writes.

Hard Drives:

• An SSD is highly recommended for all simulation work. NVMe SSDs are recommended when possible. Intermediate endurance levels are recommended (~0.3 DWPD), not QLC NAND.
• Note: For DMP, the program writes results files for every core in use, so having SSDs with a lower seek time will reduce waiting for the hard drive seeks to read/write to all those different files. It also assists with post-processing.
• Running large models out of RAM (“out-of-core”) will accelerate SSD wear (consume its write endurance) but is still preferable to not using an SSD.

Graphics:

• A mid-range graphics card (such as Nvidia Quadro P2000 and AMD Radeon Pro WX5100) can be used. These will not assist with solution time and only display graphics.
• Ansys maintains a list of tested graphics cards (other cards may/may not work as expected, but typically do work if of the same generation).
• Note: It is important to keep graphics drivers updated from the card’s vendor.
• Please refer to Ansys Help and Ansys Platform Support for a recent list of supported Graphic Card models

GPU (for solution acceleration processing):

Please see the most current "graphics cards tested" document at https://www.ansys.com/it-solutions/platform-support

Interconnect:

• If clustering, a very high-speed interconnect is mandatory (40 Gb/s plus).
• QDR or FDR IB interconnect (or better) if you will be running on two or more nodes in parallel.

Ansys Mechanical – Example Recommendations:

• Workstation with NO HPC licenses: i7-9800X (8 cores) with 64/128GB DDR4-3200
• Workstation with 1 HPC Pack: i7-9920X (12 cores) with 128GB DDR4-3200
• Workstation with 2 HPC Packs: Dual Xeon Gold 6140 (36 cores total) with 192/384 GB DDR4-2666

CPU (Processor):

• Core scaling in CFD software is very good, therefore focus on Cores * Frequency = Effective Frequency, with slight bias towards frequency.
• Both AMD and Intel CPUs are recommended, as long as the latest generation (AMD Zen cores).

Memory (RAM):

• Recommended RAM quantity is simulation dependent, but general recommendation is 4-8 GB per core, with more for small core count systems.
• Memory bandwidth is very important, therefore get high frequency RAM (i.e. 2666 MHz or faster for non-servers).
• To operate at maximum speed, all memory channels in both processors should be populated with equal amounts of memory.

Hard Drives:

• SSD not as critical but is beneficial to the user experience. High endurance SSD is not strictly necessary.
• CFD does not do a lot of Disk I/O, so you can use SATA drives.

Graphics and GPU Acceleration:

See Mechanical recommendations above, no change.

Interconnect:

• 1 Gb/s interconnect is effective for small to medium clusters, investing in better can have a good return on investment.

Ansys CFD – Example Recommendations

• Workstation with NO HPC licenses: i7-8700K (6 cores) with 64GB RAM, Ryzen 7 2700X (8 cores) with 64GB RAM
• Workstation with 1 HPC Pack: i7-9920X (12 cores) with 64/128GB RAM, Threadripper 2950X (16 cores) with 64/128GB RAM
• Workstation with 2 HPC Packs: Dual Xeon Gold 6140 (36 cores total) with 192/384 GB RAM, Dual EPYC 7351 or 7371 (32 cores total) with 128/256 GB RAM

CPU (Processor):

• Look for latest generation CPUs with high frequency.
• Note: HFSS is primarily an in-core solver, so clock speed and RAM speed is important.

Memory (RAM):

• Recommended RAM is 8 GB per core HFSS, Maxwell, and other Electronic Packages or typically 64 GB – 256 GB.
• Memory bandwidth is important, therefore get high frequency RAM (i.e. 2666 MHz).
• To operate at maximum speed, all memory channels in both processors should be populated with equal amounts of memory.

Hard Drives:

• An SSD is highly recommended for all simulation work. NVMe SSDs are recommended when possible. Intermediate endurance levels are recommended (~0.3 DWPD), not QLC NAND.

Graphics and GPU Acceleration:

See Mechanical recommendations above, no change.

Interconnect:

• QDR or FDR IB interconnect if you will be running on 2 or more nodes in parallel (on a cluster, minimum 10Gb network)

Minimum Requirements

• Detailed System Requirements
• 64-bit Intel or AMD system, running Windows 10.
• 8 GB RAM 
• A dedicated graphics card with latest drivers and at least 1GB video RAM, capable of supporting OpenGL 4.5 and DirectX 11, or higher.  Use of integrated graphics (e.g. Intel HD/IRIS) is not recommended and is not support by the Refine stage in Discovery.  See below for special graphics requirements for ANSYS Discovery Live.
• 3 button mouse

Ansys Discovery Live and Discovery Explore Stage Graphics Card Requirements

Ansys Discovery Live or the Explore stage in Ansys Discovery relies on the latest GPU technology to provide its computation and visual experience.  To run the software, you will require:

• A dedicated NVIDIA GPU card (Quadro recommended, GeForce supported) based on the Kepler, Maxwell, Pascal, or Turing architecture. Maxwell 2000 or better strongly recommended. 
• At least 4GB of video RAM (8GB recommended) on the GPU.

Minimum Requirements

64-bit Windows 7; 64-bit Windows 10; or 64-bit CentOS 7 Linux* Operating Systems

• A video card that supports OpenGL graphics
• 4 GB of free disk space
• 4 GB of RAM
• Two-button mouse with center wheel
• Screen resolution of 1280 x 1024
• * Other Linux-bases platforms are currently being tested and verified.

Optimal Rocky DEM Functionality and Basic Requirements
Recommended System Setup (for Optimal DEM Functionality)

• 8 GB of free disk space
• 8 GB of RAM
• Quad-core or better processor (Intel Core i5, Intel Core i7, or Intel Xeon processor)
• Ansys SpaceClaim or other CAD software
• Microsoft Excel or other spreadsheet software
• AVI-compatible media player

Additional Minimum Requirements for GPU or Multi-GPU Processing

One or more NVIDIA GPU cards (computing or gaming), each with the following criteria:

• At least 4 GB memory
• Fast double-precision processing capabilities*
• A CUDA compute capability of 3.5 or higher. (See the GPUs Supported table for a list of GPUs and their compute capabilities.)
• A graphics driver version** that supports the CUDA version 10 toolkit or higher. (See the CUDA Driver table for a list of which driver version supports which toolkit version.)
• * Required only for simulations using non-round particle shapes.
• **Ensure your GPU card drivers are updated before using Rocky.
• Gaming Cards | Rocky DEM:
  • GTX 1080
  • GTX 1090 Ti
  • RTX 2080
  • RTX 2080 Ti
  • Titan RTX
  • RTX 3080
  • RTX 3090
• Computing Workstation Cards | Rocky DEM:
  • Titan V
  • Quadro GP100
  • Quadro GV100
• Computing Server Cards | Rocky DEM:
  • Tesla P100
  • Tesla V100
  • Tesla A100
  • For more information, contact SimuTech Group.

Additional Requirements for CFD Coupling
You can use Rocky coupled with the following Ansys Software versions:

• Ansys version 2019 R3 - 2022R2

• Detailed System Requirements
• Ansys Optics only supports 64 bit editions of x86 processors (Intel®/AMD®) 
• ARM processors are not supported. 
• Limited support on other operating systems that are known to be compatible with the supported systems listed below such as Fedora, Rocky Linux, and OpenSUSE.
• Ansys Optics Finite Element IDE (FE IDE) for; CHARGE, HEAT, DGTD, FEEM, MQW simulation tools, uses, a 3D rendering and visualization library that requires a compliant implementation of OpenGL 3.2+ or DirectX 11. For optimal performance, please install the FE IDE directly on a locally accessible host computer with adequate graphics resources and up-to-date graphics drivers. The FE IDE has limited support on systems without compliant graphics drivers, including virtual machines and remote desktop interfaces. 

Supported Operating Systems

• Windows10 and newer (Professional, Enterprise & Education)
• Windows Server 2016 and newer (Standard Edition)
• RedHat Enterprise 7.9 to latest
• RedHat Enterprise 8.7 to latest
• CentOS 7.9 to 8.3 
• SUSE Linux Enterprise Server & Desktop 12 SP5 & 15 SP5
• Ubuntu 20.04 and 22.04 

Supported Technologies

• Amazon EC2
• Microsoft Azure
• HPC network technologies including: Infiniband, Myrinet, Gb and 10Gb Ethernet
• MPI frameworks: Microsoft MPI, MPICH/MPICH2, OpenMPI, IntelMPI and other
  commercial MPI distributions
• HPC schedulers: SLURM, SGE, PBS Pro, OpenPBS, Platform LSF, Torque, and HTCondor

See the Ansys Platform Support page for full platform support information for the current Ansys release, including certification and test information for operating systems, graphics cards, job schedulers and other platform components.   

Hardware specifications guidelines

Memory Size (RAM)

The memory size determines the size of the simulation design or project that can run on the computer. It does not affect the simulation speed as long as the entire simulation can fit into the RAM, otherwise the computer may be forced to swap memory to the hard drive (or the application will provide an error). If swapping occurs, the simulation speed will be extremely slow.

Desktop computers nowadays typically have 8-32 GB of RAM, which is sufficient for running a large fraction of simulations. Workstations often have 32-128 GB of RAM, which is sufficient for running almost all simulations. You can check how much memory is required to run FDTD simulation.

Memory Bandwidth (RAM)

When running Lumerical simulation, large amounts of data must be continuously transferred between the RAM and CPU. When the memory bus is unable to transfer data fast enough, the processor is forced to wait, limiting the overall speed of your simulation. For example, on a typical desktop computer with 8 cores, the simulation speed might increase by a factor of 2-3x when using eight cores compared to one core. Therefore, when purchasing a computer, having a fast memory bandwidth is very important.

CPU Core Count

Our simulation tools will try and use all of your CPU cores to run as quickly as possible. However, as we learned above, our simulation is memory bandwidth limited, so adding more cores does not always make the simulation go faster. The speedup that most users see when moving to higher core count CPUs is most often due to other (memory related) improvements to the CPU architecture.

Most CPUs’ support hyper-threading, which allows the operating system to treat each physical CPU core as two logical cores. This feature does not provide any speed increase for FDTD simulations because the overall performance bottleneck is the data transfer rate between the CPU and RAM, not the actual computing capability of the cores.

CPU Clock Speed

The CPU clock speed is typically not the most important factor for Lumerical simulation speeds. While a faster clock speed does allow each core to run more quickly, the overall simulation speed is limited by the access between CPU and RAM.

Workstations With Multiple CPU’s

Workstations with multiple CPU’s are a good way to increase the simulation speed. The most important factor is that each CPU has its own memory bus connection to the RAM. As explained above, the data transfer rate between CPU and RAM is the performance bottleneck, so having one memory bus per CPU allows the simulation speed scale very well with the number of CPU’s.

Example:

The Intel Xeon Gold 5115 has a ‘Maximum memory bandwidth’ of 107 GB/s. Up to 4 of these processors can be installed in a single workstation, which would give a total bandwidth of 4 x 107 = 430 GB/s. To achieve optimum performance, DDR4-2400 memory modules should be used.

References:
- Intel Xeon Gold Processor 65
- Xeon Gold 5115 Processor Specs 63
- Intel Xeon Processors - Value Analysis 65

Clusters (Multiple Computers on a Network)

In applications where a single computer is not enough, multiple computers can be connected over a network to form a “cluster”. FDTD Solutions supports 2 modes when running on a cluster: running multiple simulations across a network (Concurrent Parametric Computing) and running a single, large complex simulation across multiple computers (Distributed Computing).

Network Speed and Latency

• When running a simulation locally on a single computer, the network speed does not have any effect on the simulation speed.
• When running a simulation remotely on a single computer, or running multiple simulations on a cluster  (ie. a sweep or optimization), network latency has no effect on the simulation speed, and network speed will only affect how fast results can be retrieved.
• When running a single, large simulation across multiple computers , the network speed is extremely important - high speed, low latency interconnects such as InfiniBand are recommended in such cases.

Cloud Computing

You can use cloud computing services to evaluate the latest hardware before making a purchasing decision, or in cases where a long term hardware investment does not make sense:

• Amazon Web Services: Instance Types  “General Purpose” and “Compute Optimized”. Running FDTD Solutions on AWS .
• Microsoft Azure: Instance Types:  “General Purpose”, “Compute Optimized”, and “High Performance Computing”. Running FDTD Solutions on Microsoft Azure .
• more...

GPU support

Starting with the 2023 R2 release, Lumerical FDTD now supports GPU processing. See this Knowledge Base (KB) for more information. 

Notes: All the above examples are not intended to be endorsements of these models or brands. They are simply examples used to illustrate the points described in the page.