• 1. Introduction
• 2. Choosing a Virtual Machine
  • 2.1. Categories of Instance Types
  • 2.2. Understanding Instance Type Names
  • 2.3. Simultaneous Multithreading/Hyper-Threading
  • 2.4. CPUs/vCPUs vs Memory
• 3. NVIDIA GPU Driver Support
  • 3.1. NVIDIA GPU Driver Types
  • 3.2. NVIDIA GPU Driver Support by Application
  • 3.3. NVIDIA GPU Driver Support by Instance Type
  • 3.4. Configuring Slurm Clusters with NVIDIA GPU Instances
• 4. Application Support
  • 4.1. Ansys Applications
    • 4.1.1. Ansys 2024 R2 Applications
    • 4.1.2. Ansys 2024 R1 Applications
    • 4.1.3. Ansys 2023 R2 Applications
    • 4.1.4. Ansys 2023 R1 Applications
    • 4.1.5. Ansys Service Pack Installation
  • 4.2. Third-Party Applications
• 5. Recommended Hardware for VDI Workflows
  • 5.1. General Tips and Guidelines for Choosing VDI Hardware
  • 5.2. Recommended Instance Types for VDI Workflows
• 6. Recommended Hardware for HPC Workflows
  • 6.1. General Tips and Guidelines for Choosing HPC Hardware
  • 6.2. Recommended Instance Types for Cluster Workflows
  • 6.3. Elastic Fabric Adapter (EFA)
• 7. Requirements and Best Practices for Slurm Autoscaling Clusters
  • 7.1. Overview of Slurm Autoscaling Cluster Setups
  • 7.2. Best Practices for Slurm Autoscaling Clusters
    • 7.2.1. Best Practices when Creating an Autoscaling Cluster
    • 7.2.2. Best Practices When Adding Ansys HPC Platform Services (HPS) to an Autoscaling Cluster
    • 7.2.3. Best Practices for Using a Slurm Autoscaling Cluster
  • 7.3. Testing a Slurm Autoscaling Cluster
    • 7.3.1. Check the Readiness of the Slurm Autoscaling Cluster
    • 7.3.2. Submit a Test Job to the Slurm Autoscaling Cluster
• 8. Requirements and Best Practices for Slurm HPC Clusters
  • 8.1. Overview of Slurm HPC Cluster Setups
  • 8.2. Configuring Resources for a Slurm HPC Cluster Workflow
  • 8.3. Testing a Slurm HPC Cluster
    • 8.3.1. Check the Readiness of the Slurm HPC Cluster
    • 8.3.2. Submit a Test Job to the Slurm HPC Cluster
  • 8.4. Best Practices and Limitations for Slurm HPC Clusters
  • 8.5. Workaround for Reusing a Slurm Controller
• 9. Recommended Configurations by Application
  • 9.1. Ansys Discovery Virtual Desktop
  • 9.2. Ansys Electronics Desktop Configurations
    • 9.2.1. AEDT Virtual Desktop
    • 9.2.2. AEDT Autoscaling Cluster Workflow
      • 9.2.2.1. Step 1: Create an AEDT Autoscaling Cluster
      • 9.2.2.2. (Optional) Step 2: Create a Virtual Desktop for AEDT Job Submission
      • 9.2.2.3. Step 3: Configure Job Submission to HPC Platform Services
      • 9.2.2.4. Step 4: Submit a Job to the AEDT Autoscaling Cluster
    • 9.2.3. AEDT HPC Cluster (2024 R1 or 2023 R2)
      • 9.2.3.1. Step 1: Create Shared NFS Storage with the AEDT Installation
      • 9.2.3.2. Step 2: Create an AEDT Cluster
      • 9.2.3.3. Step 3: Connect to the AEDT Cluster from an AEDT Virtual Desktop
      • 9.2.3.4. Step 4: Submit a Job to the AEDT Cluster
      • 9.2.3.5. Step 5: Monitor the Job
      • 9.2.3.6. Step 6: Retrieve Output Files
      • 9.2.3.7. Limitations
    • 9.2.4. AEDT HPC Cluster (2023 R1)
      • 9.2.4.1. Step 1: Prepare an AEDT Virtual Desktop for Job Submission
      • 9.2.4.2. Step 2: Create an AEDT Cluster
      • 9.2.4.3. Step 3: Connect to the AEDT Cluster
      • 9.2.4.4. Step 4: Submit a Job to the AEDT Cluster
      • 9.2.4.5. Step 5: Monitor the Job
      • 9.2.4.6. Step 6: Retrieve Output Files
  • 9.3. Ansys EMA3D Virtual Desktop
  • 9.4. Ansys Fluids Configurations
    • 9.4.1. Virtual Desktop with Ansys Fluids
      • 9.4.1.1. Creating a Virtual Desktop with Ansys Fluids
      • 9.4.1.2. Running the Fluent GPU Solver on a Virtual Desktop
      • 9.4.1.3. Running Fluent 2023 R1 on Windows on hpc6a or hpc7a
      • 9.4.1.4. Running Fluent Aero and Fluent Icing (2023 R2 and Later)
      • 9.4.1.5. Requirements for EnSight Remote Launcher
    • 9.4.2. Ansys Fluids Autoscaling Cluster
      • 9.4.2.1. Step 1: Create an Ansys Fluids Autoscaling Cluster
      • 9.4.2.2. Step 2: Create an Ansys Fluids Virtual Desktop for Job Submission
      • 9.4.2.3. Step 3: Copy Input Files to the Shared Storage on the Virtual Desktop
      • 9.4.2.4. Step 4: Submit a Job to the Ansys Fluids Autoscaling Cluster
        • 9.4.2.4.1. Running Ansys Fluent (CPU Compute)
          • 9.4.2.4.1.1. Run Fluent Interactively with Slurm and Use Cluster Nodes (CPU Compute)
          • 9.4.2.4.1.2. Run Fluent with Slurm in Batch Mode Using the Command Line (CPU Compute)
          • 9.4.2.4.1.3. Run Fluent in Batch Mode Using a Slurm Script (CPU Compute)
        • 9.4.2.4.2. Running Ansys CFX
          • 9.4.2.4.2.1. Run the CFX Solver Manager interactively with Slurm and use cluster nodes
          • 9.4.2.4.2.2. Run CFX with Slurm in Batch Mode Using the Command line
          • 9.4.2.4.2.3. Run CFX in Batch Mode Using a Slurm Script
        • 9.4.2.4.3. Ansys Fluent and CFX: Workarounds for Intel MPI-Related Issues
    • 9.4.3. Ansys Fluids HPC Cluster (2024 R1 or Earlier)
      • 9.4.3.1. Creating Resources for a Fluids HPC Cluster Workflow
        • 9.4.3.1.1. Step 1: Create a Virtual Machine with Shared NFS Storage
        • 9.4.3.1.2. Step 2: Create a Virtual Desktop with Slurm Controller and KDE
        • 9.4.3.1.3. Step 3: Add Ansys Fluids to the Virtual Desktop
        • 9.4.3.1.4. Step 4: Create the Slurm Cluster
      • 9.4.3.2. Important Notes about the Fluids HPC Cluster Workflow
      • 9.4.3.3. Transferring Files to/from the Slurm Controller Virtual Desktop
      • 9.4.3.4. Submitting Jobs to a Fluids HPC Cluster
        • 9.4.3.4.1. Running Ansys Fluent (CPU Compute)
          • 9.4.3.4.1.1. Run small Fluent jobs on the Slurm Controller machine
          • 9.4.3.4.1.2. Run Fluent interactively with Slurm and use cluster nodes
          • 9.4.3.4.1.3. Run Fluent with Slurm in batch mode using the command line
          • 9.4.3.4.1.4. Run Fluent in batch mode using a Slurm script
        • 9.4.3.4.2. Running Fluent (GPU Compute)
          • 9.4.3.4.2.1. Run Fluent with Slurm in batch mode using the command line
          • 9.4.3.4.2.2. Run Fluent in batch mode using a Slurm script
          • 9.4.3.4.2.3. Run Fluent interactively with Slurm and use cluster nodes (requires workaround)
        • 9.4.3.4.3. Running Ansys CFX
          • 9.4.3.4.3.1. Run small CFX jobs on the Slurm Controller machine
          • 9.4.3.4.3.2. Run the CFX Solver Manager interactively with Slurm and use cluster nodes
          • 9.4.3.4.3.3. Run CFX with Slurm in batch mode using the command line
      • 9.4.3.5. Resizing the HPC Cluster
      • 9.4.3.6. Optional: Connecting to the Slurm Controller Machine via SSH
  • 9.5. Ansys Granta MI Pro Configuration
    • 9.5.1. Deployment Overview
    • 9.5.2. Steps for Configuring an Ansys Granta MI Pro Virtual Desktop
      • 9.5.2.1. Step 1: Create a New Virtual Desktop
      • 9.5.2.2. Step 2: Restart the Virtual Desktop
      • 9.5.2.3. Step 3: Install Prerequisite Software for Ansys Material Calibration
      • 9.5.2.4. Step 4: Add the Granta MI Pro Application
      • 9.5.2.5. Step 5: Verify the Granta MI Pro Installation
      • 9.5.2.6. Step 6: Add MI Pro System Users
      • 9.5.2.7. Step 7: Install the Granta MI Materials Gateway client software
  • 9.6. Ansys Granta Selector Virtual Desktop
  • 9.7. Ansys LS-DYNA Configurations
    • 9.7.1. Ansys LS-DYNA Virtual Desktop
      • 9.7.1.1. Create a Virtual Desktop with LS-DYNA
      • 9.7.1.2. Run a Windows LS-DYNA Virtual Desktop
      • 9.7.1.3. Run a Linux LS-DYNA Virtual Desktop
        • 9.7.1.3.1. Run a Linux LS-DYNA Virtual Desktop Using SSH
        • 9.7.1.3.2. Run a Linux LS-DYNA Virtual Desktop Using GNOME Desktop
    • 9.7.2. Ansys LS-DYNA Autoscaling Cluster Workflow
      • 9.7.2.1. Step 1: Create an LS-DYNA Autoscaling Cluster
      • 9.7.2.2. Step 2: Create an LS-DYNA Virtual Desktop for Job Submission
      • 9.7.2.3. Step 3: Copy Input Files to the Shared Storage on the Virtual Desktop
      • 9.7.2.4. Step 4: Submit a Job to the LS-DYNA Autoscaling Cluster
    • 9.7.3. LS-DYNA HPC Cluster with EFA (2024 R1 or Earlier)
      • 9.7.3.1. Step 1: Create a Virtual Machine with Shared NFS Storage
      • 9.7.3.2. Step 2: Create a Virtual Desktop with Slurm Controller
      • 9.7.3.3. Step 3: Create an LS-DYNA EFA Cluster
      • 9.7.3.4. Step 4: Connect to the Controller and Launch a Job
      • 9.7.3.5. Postprocessing
  • 9.8. Ansys Lumerical Configurations
    • 9.8.1. Ansys Lumerical Virtual Desktop
    • 9.8.2. Ansys Lumerical Autoscaling Cluster Workflow
      • 9.8.2.1. Step 1: Create an Ansys Lumerical Autoscaling Cluster
      • 9.8.2.2. Step 2: Submit Jobs to the Ansys Lumerical Autoscaling Cluster
        • 9.8.2.2.1. Submit Jobs to the Cluster Using Ansys Lumerical FDTD
        • 9.8.2.2.2. Submit Jobs to the Cluster Using the lumslurm API
  • 9.9. Ansys Mechanical Configurations
    • 9.9.1. Windows Virtual Desktop with Ansys Mechanical Enterprise
    • 9.9.2. Ansys Mechanical Autoscaling Cluster Workflow
      • 9.9.2.1. Step 1: Create an Ansys Structures Autoscaling Cluster
      • 9.9.2.2. (Optional) Step 2: Create a Virtual Desktop for Mechanical Job Submission
      • 9.9.2.3. Step 3: Submit a Job to the Ansys Structures Autoscaling Cluster
        • 9.9.2.3.1. Specify a Default Queue for Mechanical Jobs
        • 9.9.2.3.2. Submit a Mechanical Job to HPC Platform Services from Mechanical
        • 9.9.2.3.3. Submit a Mechanical Job to HPC Platform Services from Workbench
        • 9.9.2.3.4. Submit a Mechanical Job to HPC Platform Services using the HPC Manager Web App
        • 9.9.2.3.5. Submit a Mechanical Job to HPC Platform Services Using Python
    • 9.9.3. Mechanical Ansys RSM Cluster (2024 R1 or Earlier)
      • 9.9.3.1. (Optional) Step 1: Create a Virtual Machine with Shared File Storage
      • 9.9.3.2. Step 2: Create an HPC Cluster
      • 9.9.3.3. Step 3: Connect to the Cluster
      • 9.9.3.4. Step 4: Set Up a Mechanical Solution
      • 9.9.3.5. Step 5: Get Mechanical Results
    • 9.9.4. DCS Cluster for Workbench Mechanical (2024 R1 or Earlier)
      • 9.9.4.1. Step 1: Create a Virtual Desktop with the DCS Design Point Service
      • 9.9.4.2. (Optional) Step 2: Create a Virtual Machine with Shared NFS Storage
      • 9.9.4.3. Step 3: Create a DCS Cluster/Evaluator
      • 9.9.4.4. Step 4: Use the DCS for Design Points Service
      • 9.9.4.5. Automatic Reconfiguration of DCS Evaluators
  • 9.10. Ansys Medini Virtual Desktop
  • 9.11. Ansys ModelCenter Virtual Desktop Configurations
  • 9.12. Ansys Motion Virtual Desktop
  • 9.13. Ansys Motor-CAD Virtual Desktop
  • 9.14. Ansys nCode Virtual Desktop
  • 9.15. Ansys optiSLang Configurations
    • 9.15.1. Connector Support in optiSLang Workflows
    • 9.15.2. Ansys optiSLang Virtual Desktop
    • 9.15.3. Ansys optiSLang Workflow with an Ansys Mechanical DCS Cluster (2024 R1 or Earlier)
      • 9.15.3.1. Set Up Resources for the optiSLang/DCS Workflow
        • 9.15.3.1.1. Step 1: Set up an Ansys Mechanical DCS Cluster
        • 9.15.3.1.2. Step 2: Set up a Virtual Desktop with Ansys optiSLang and Mechanical Enterprise
      • 9.15.3.2. Connect to the optiSLang Virtual Desktop
      • 9.15.3.3. Set up an optiSLang/DCS Workflow
        • 9.15.3.3.1. Step 1: Sync a Workbench Project with DCS
        • 9.15.3.3.2. Step 2: Set up a Workflow in optiSLang
        • 9.15.3.3.3. Step 3: Run the Workflow
  • 9.16. Ansys PathFinder-SC Autoscaling Cluster Workflow
    • 9.16.1. Step 1: Create an Ansys Pathfinder-SC Autoscaling Cluster
    • 9.16.2. Step 2: Create a Virtual Desktop for Job Submission
    • 9.16.3. Step 3: Connect to the Virtual Desktop and Launch Workers
  • 9.17. Ansys RaptorH/RaptorX/VeloceRF Virtual Desktop
    • 9.17.1. Create a Virtual Desktop with RaptorH/RaptorX/VeloceRF
    • 9.17.2. Connect to the Virtual Desktop via KDE
    • 9.17.3. (Optional) Connect to the Virtual Desktop via SSH
  • 9.18. Ansys RedHawk-SC Configurations
    • 9.18.1. Ansys RedHawk-SC Autoscaling Cluster Workflow
      • 9.18.1.1. Step 1: Create an Ansys RedHawk-SC Autoscaling Cluster
      • 9.18.1.2. Step 2: Create a Virtual Desktop for Job Submission
      • 9.18.1.3. Step 3: Connect to the Virtual Desktop and Launch Workers
    • 9.18.2. Ansys RedHawk-SC HPC Cluster Workflow (2024 R1 or Earlier)
      • 9.18.2.1. General Guidelines for RedHawk-SC Workflows
      • 9.18.2.2. Virtual Machine with Shared NFS Storage
      • 9.18.2.3. RedHawk-SC HPC Cluster
        • 9.18.2.3.1. Step 1: Create a Virtual Desktop with Slurm Controller and Ansys Semiconductor CentOS 7
        • 9.18.2.3.2. Step 2: Create the Slurm Cluster for RedHawk-SC Workers
        • 9.18.2.3.3. Step 3: Prepare a Virtual Desktop for Simulation Work
        • 9.18.2.3.4. Transferring Files to/from the Slurm Controller Virtual Desktop
        • 9.18.2.3.5. Resizing the Cluster
        • 9.18.2.3.6. Optional: Connecting to the Slurm Controller Machine via SSH
      • 9.18.2.4. Virtual Desktop Slurm Node with Ansys RedHawk-SC
  • 9.19. Ansys Rocky Virtual Desktop
  • 9.20. Ansys SCADE Virtual Desktop
  • 9.21. Ansys Speos Configurations
    • 9.21.1. Ansys Speos Virtual Desktop
    • 9.21.2. Ansys Speos Autoscaling Cluster Workflow
      • 9.21.2.1. Step 1: Create an Ansys Speos Autoscaling Cluster
      • 9.21.2.2. Step 2: Configure SpeosCore and Batch Scripts
      • 9.21.2.3. Step 3: Submit a Job to the Speos Autoscaling Cluster
    • 9.21.3. Ansys Speos HPC Cluster (2024 R1 or Earlier)
      • 9.21.3.1. Step 1: Create a Virtual Machine with Shared NFS Storage
      • 9.21.3.2. Step 2: Install the Slurm Controller
      • 9.21.3.3. Step 3: Install Speos HPC
      • 9.21.3.4. Step 4: Create an HPC Cluster
      • 9.21.3.5. Step 5: Configure SpeosCore and Batch Scripts
  • 9.22. Ansys Totem-SC Configurations
    • 9.22.1. Ansys Totem-SC Autoscaling Cluster Workflow
      • 9.22.1.1. Step 1: Create an Ansys Totem-SC Autoscaling Cluster
      • 9.22.1.2. Step 2: Create a Virtual Desktop for Job Submission
      • 9.22.1.3. Step 3: Connect to the Virtual Desktop and Launch Workers
    • 9.22.2. Ansys Totem-SC HPC Cluster Workflow (2024 R1 and Earlier)
      • 9.22.2.1. General Guidelines for Ansys Totem-SC Configurations
      • 9.22.2.2. Virtual Machine with Shared NFS Storage
      • 9.22.2.3. Totem-SC HPC Cluster
        • 9.22.2.3.1. Step 1: Create a Virtual Desktop with Slurm Controller and Ansys Semiconductor CentOS 7
        • 9.22.2.3.2. Step 2: Create the Slurm Cluster
        • 9.22.2.3.3. Step 3: Prepare a Virtual Desktop for Simulation Work
        • 9.22.2.3.4. Transferring Files to/from the Slurm Controller Virtual Desktop
        • 9.22.2.3.5. Resizing the Cluster
      • 9.22.2.4. Virtual Desktop Slurm Node with Ansys Totem-SC
  • 9.23. Ansys Workbench LS-DYNA Configurations
    • 9.23.1. Ansys Workbench LS-DYNA Autoscaling Cluster Workflow via HPC Platform Services
      • 9.23.1.1. Step 1: Create an Ansys Structures Autoscaling Cluster
      • 9.23.1.2. (Optional) Step 2: Create a Virtual Desktop for Workbench LS-DYNA Job Submission
      • 9.23.1.3. Step 3: Submit a Workbench LS-DYNA Job to the Ansys Structures Autoscaling Cluster
        • 9.23.1.3.1. Submit a Workbench LS-DYNA Job to HPC Platform Services from Mechanical
        • 9.23.1.3.2. Submit a Workbench LS-DYNA Job Using the HPC Manager Web App
  • 9.24. Ansys Zemax OpticStudio Virtual Desktop
• 10. Ansys HPC Platform Services
  • 10.1. Overview of Ansys HPC Platform Services
  • 10.2. Integrating Ansys HPC Platform Services with Autoscaling Clusters
  • 10.3. Do I Need a Docker Account to Install HPC Platform Services?
  • 10.4. Changing the Default Credentials for Accessing Ansys HPC Manager
  • 10.5. Specifying a Default Queue for an Autoscaling Cluster (HPS Workflows)
  • 10.6. Specifying Default Queues for Cluster Applications (HPS Workflows)
  • 10.7. Submitting Jobs to Ansys HPC Platform Services
  • 10.8. Launching the Ansys HPC Manager Web App
  • 10.9. Selecting a Queue when Submitting a Job to Ansys HPC Platform Services
  • 10.10. Downloading Result Files from Ansys HPC Manager
  • 10.11. Installing PyHPS for Job Submission to Ansys HPC Platform Services Using Python
• 11. Best Practices for Managing Costs
  • 11.1. Best Practices for Project Spaces
  • 11.2. Best Practices for Resources
  • 11.3. Best Practices for Tenant Administrators