Run Ansys Discovery on Azure for simulation-driven design

This article describes a reference architecture for running Ansys Discovery on Azure NVads V710 v5 virtual machines, including validated test results across computational fluid dynamics (CFD), thermal, and structural workloads.

Overview

Ansys Discovery is an interactive engineering application that unifies 3D geometry modeling, real-time simulation, and visual analysis into a single workspace. It accelerates early-stage engineering exploration and maintains continuity into high-fidelity validation workflows.

Discovery enables engineers and designers to create, modify, simulate, and iterate on designs in real time. This approach reduces the time required to assess design feasibility, physics behavior, and performance trade-offs. Traditional workflows treat computer-aided design (CAD) and simulation as sequential steps. Discovery collapses these steps into a continuous feedback loop that supports faster decision-making and more informed design outcomes.

This architecture is intended for design engineers, simulation-aware designers, and product engineers who participate in earlier analysis cycles.

Challenges

Globally distributed engineering teams often depend on high-end local workstations and separate CAD and simulation tools, which lengthens design cycles and delays validation. Provisioning and refreshing GPU workstations for each engineer is costly and slow, and it leaves expensive hardware underused between peak design cycles. When CAD models and simulation results are scattered across individual machines, collaboration suffers, data is duplicated, and traceability from concept to validated design breaks down. Because early-stage analysis depends on switching between disconnected tools, engineers wait for simulation results before they can iterate, which delays feasibility decisions. Keeping proprietary designs on local workstations also increases the risk of intellectual property exposure and complicates access control.

Core capabilities

Ansys Discovery is organized into three stages that you can switch between freely:

• Model: Direct 3D geometry creation and modification, without simulation overhead.
• Explore: Instant, real-time simulation results that use GPU-accelerated meshing and solvers for rapid insight.
• Refine: Higher-fidelity simulation that uses body-fitted meshing and CPU and GPU solvers for design validation.

Key benefits

Running Ansys Discovery on Microsoft Azure provides a scalable, cloud-based platform for simulation-driven design with the following advantages:

• Anywhere access for distributed teams: Engineers run Discovery on Azure GPU-enabled virtual machines from any location, without local GPUs or hardware refresh cycles.
• Fractional GPU allocation and elastic scaling: Azure NVads V710 v5 supports fractional GPU allocation (1/6 to full GPU), and Azure provides rapid provisioning to match peak design cycles.
• Real-time interactive performance: Discovery's GPU-accelerated Explore mode uses Azure GPU infrastructure to deliver instant physics feedback and faster design iteration.
• Centralized storage and design traceability: Azure Files or Azure NetApp Files store CAD models, simulation results, and design artifacts, which reduces duplication and maintains traceability from concept through validated design.
• Intellectual property protection: Designs remain in enterprise-managed Azure environments with access controlled by Microsoft Entra ID and Azure security services.

Architecture

The Ansys Discovery on Azure reference architecture supports interactive, GPU-accelerated simulation-driven design on scalable Azure infrastructure. It supports early-stage engineering exploration, rapid iteration, and design optimization. It integrates with downstream high-fidelity simulation, product lifecycle management (PLM), and digital thread systems.

Components

The reference architecture for Ansys Discovery on Azure includes the following Azure components:

• Azure Virtual Machines (GPU-enabled): Provides the compute platform that runs Discovery. NVads V710 v5 instances deliver AMD Radeon Pro V710 GPU acceleration for interactive simulation workflows. For higher-fidelity Refine-stage validation, you can pair the GPU VM with a CPU-optimized VM for body-fitted meshing and advanced solver workloads.

• Ansys license server (Azure Virtual Machine): A dedicated VM in the application subnet that runs the Ansys licensing service. Discovery VMs check out licenses from this server over the virtual network.

• GPU drivers (AMD Radeon Pro V710): Required to enable GPU acceleration for Discovery workloads on Azure VMs.

• Azure Virtual Network with accelerated networking: Provides secure, low-latency connectivity between users, the VM, and storage, with improved throughput for graphics and data transfer. The architecture uses a hub-and-spoke topology, where the workload runs in a spoke virtual network that connects to a shared hub through virtual network peering.

• Azure Firewall: Provides centralized network traffic filtering and threat protection for traffic that enters and leaves the hub virtual network.

• Azure Bastion: Provides secure RDP and SSH access to the VMs without exposing public IP addresses on the workload.

• Azure VPN Gateway: Provides encrypted connectivity between on-premises engineering networks and the hub virtual network. You can use Azure ExpressRoute instead for private, high-bandwidth connectivity.

• Azure Files or Azure NetApp Files: Provides shared storage for CAD models, simulation data, user profiles, and project artifacts, removing the dependency on local storage.

• Microsoft Entra ID: Controls user sign-in to the VM and enforces role-based access control (RBAC) for the engineering team.

• Azure Key Vault: Stores and manages secrets, keys, and certificates that the solution uses, such as credentials and encryption keys.

Workflow

The following workflow corresponds to the numbered steps in the preceding diagram. It combines the network access path with the engineer's design activities:

1. A simulation user signs in to Ansys Discovery on an Azure-hosted virtual workstation over the internet.
2. Microsoft Entra ID authenticates the user and applies role-based access control (RBAC) before it grants access to the workload.
3. Traffic enters the hub virtual network, where Azure Firewall filters inbound and outbound traffic, Azure Bastion provides secure RDP and SSH access, and Azure VPN Gateway terminates encrypted connections.
4. Authenticated HTTPS traffic routes from the hub to the spoke virtual network through virtual network peering.
5. In the application subnet, the NVads V710 v5 GPU virtual machine checks out a license from the Ansys license server. The engineer imports CAD geometry into Discovery and begins modification and exploration. Azure GPU resources deliver real-time simulation and immediate physics feedback, and the engineer refines the design interactively and adjusts fidelity controls to balance speed and accuracy.
6. Discovery reads and writes CAD models, simulation results, and project artifacts on Azure NetApp Files or Azure Files in the storage subnet over SMB or NFS.
7. On-premises engineering and support teams connect to the environment through Azure ExpressRoute or a site-to-site VPN. This connectivity enables hybrid access and handoff of validated designs to downstream solvers or enterprise product lifecycle management (PLM) systems for further analysis and lifecycle management.

Recommended VM sizes

Ansys Discovery on Azure runs on the Azure NVads V710 v5 series. These VMs are designed for graphics-intensive, interactive workloads and map directly to Discovery's real-time, GPU-accelerated Explore mode, where instant feedback matters more than maximum solver throughput. They suit Explore-mode activities including concept-level CFD, structural, and thermal studies; early-stage topology and shape experimentation; and design review sessions with rapid iteration loops.

NVads V710 v5 VMs provide the following capabilities:

• AMD Radeon Pro V710 GPUs
• Fractional GPU support (1/6 to full GPU) for cost-optimized right-sizing
• High-frame-rate remote graphics for interactive visualization
• No additional GPU licensing required

You can match the VM size to model complexity and concurrency needs as listed in the following table:

The fractional GPU model lets organizations right-size cost and performance based on user needs, from individual engineers to power users. This model provides flexibility and maintains high-quality visualization and responsiveness.

Scenario details

A manufacturing organization adopts Ansys Discovery on Azure NVads V710 v5 virtual workstations to enable simulation-driven design for globally distributed engineering teams. Traditionally, these teams relied on high-end local workstations and separate CAD and simulation tools, which resulted in long design cycles and delayed validation.

By running Discovery on Azure, engineers create, modify, simulate, and optimize products within a unified environment while they use Azure-hosted AMD Radeon Pro V710 GPUs for real-time simulation and visualization. A single environment supports the complete design workflow across the Model, Explore, and Refine stages, backed by Azure NVads V710 v5 VMs, Azure Files or Azure NetApp Files for centralized storage, Microsoft Entra ID for secure access, and Azure networking for low-latency remote visualization.

Potential use cases

Ansys Discovery on Azure enables real-time, simulation-driven design for fluid, thermal, structural, electromagnetic, and optimization workloads. It helps engineering teams accelerate product development, reduce prototyping costs, and improve design quality through cloud-based GPU acceleration. Consider this architecture for the following use cases:

• Real-time design exploration and rapid product prototyping.
• Computational fluid dynamics (CFD) analysis for aerodynamics, internal flows, valves, pumps, and heating, ventilation, and air conditioning (HVAC) systems.
• Thermal management for electronics, batteries, power devices, and cooling systems.
• Structural validation and stress analysis of components and assemblies.
• Modal and vibration analysis to identify resonance and durability issues.
• Topology optimization for lightweight and performance-driven designs.
• Heat exchanger and fluid-thermal system optimization.
• Antenna and electromagnetic design validation.
• Cloud-based engineering workstations for distributed design teams.
• Simulation-driven design workflows across the manufacturing, automotive, aerospace, electronics, and industrial equipment industries.

Considerations

The following considerations align with the pillars of the Azure Well-Architected Framework, which is a set of guiding tenets used to improve the quality of a workload. For more information, see Well-Architected Framework.

Reliability

Reliability helps ensure that your application can meet the commitments that you make to your customers. For more information, see Design review checklist for Reliability.

• Consider availability sets or availability zones for production environments.
• Use Azure storage redundancy options to protect design and simulation data.
• Implement backup strategies for CAD and simulation artifacts.

Security

Security provides protections against deliberate attacks and the misuse of your valuable data and systems. For more information, see Design review checklist for Security.

• Enforce authentication through Microsoft Entra ID with role-based access control (RBAC).
• Use network security groups (NSGs) to restrict traffic to authorized users and services.
• Store secrets, keys, and certificates in Azure Key Vault rather than in application configuration or on the VMs.
• Protect sensitive engineering data by keeping it within Azure-managed storage and preventing local data sprawl.

Cost optimization

Cost Optimization focuses on ways to reduce unnecessary expenses and improve operational efficiencies. For more information, see Design review checklist for Cost Optimization.

AMD Radeon Pro V710 GPUs on Azure reduce costs through fractional GPU allocation, on-demand scaling, and the elimination of expensive dedicated engineering workstations. This approach lets organizations align GPU spending with actual engineering workload requirements.

Azure NVads V710 v5 enables cost-efficient GPU consumption through fractional GPU allocation that ranges from 1/6 GPU to a full GPU. Organizations can match infrastructure sizing to user requirements, reduce idle GPU capacity, avoid workstation refresh cycles, and scale resources on demand during peak engineering periods. This approach lowers total cost of ownership while it maintains high-performance graphics and simulation capabilities.

Operational excellence

Operational Excellence covers the operations processes that deploy an application and keep it running in production. For more information, see Design review checklist for Operational Excellence.

Azure enables centralized deployment, management, monitoring, and scaling of Discovery environments. Organizations can standardize engineering workstations, simplify software management, reduce IT overhead, and provide consistent user experiences across global teams while they maintain centralized governance and security controls.

Ansys Discovery on Azure improves operational excellence by providing a scalable, centrally managed simulation environment that lets engineers access high-performance GPU workstations from anywhere. Azure simplifies the deployment, management, and scaling of engineering resources while it reduces dependency on costly local hardware. Centralized storage, security controls through Microsoft Entra ID, and cloud-based collaboration help organizations streamline engineering operations, support globally distributed teams, accelerate design cycles, and improve resource usage through flexible GPU allocation and on-demand infrastructure.

Performance efficiency

Performance Efficiency refers to your workload's ability to scale to meet the demands that users place on it efficiently. For more information, see Design review checklist for Performance Efficiency.

• Use GPU-enabled virtual machines (NVads V710 v5) to optimize performance for real-time simulation workloads.
• Match VM size and fractional GPU allocation to workload complexity (Explore mode for interactive work, Refine mode for validation).
• Ensure low-latency network connectivity between users and Azure regions to maintain interactive performance.
• Scale VM sizes up or down based on model size and simulation complexity, and use multiple VMs to support distributed design teams.
• Use Azure elasticity to support burst usage during peak engineering cycles.

Validation results

Microsoft validated Discovery on Azure NVads V710 v5 across CFD, thermal, structural, and modal scenarios. All workloads passed, which demonstrates consistent performance across simulation types. Solve times stay fast even at multi-million element scale, which supports the use of Discovery for real-time engineering exploration at enterprise scale.

Test configuration

The following results were captured on this tested configuration:

Test scenarios

The following test scenarios cover CFD, thermal, structural, and modal simulation types.

Test results

Azure NVads V710 v5 is optimized for Discovery's real-time simulation workflows. It provides GPU acceleration tuned for interactive engineering rather than batch compute alone. The fractional GPU model lets organizations right-size cost and performance based on user needs, from individual engineers to power users. This approach provides flexibility while it maintains high-quality visualization and responsiveness.

Deploy this scenario

Prerequisites

Before you deploy this architecture, make sure you have the following items:

• An active Azure subscription with quota for the NVads V710 v5 series in your target region.
• A valid Ansys Discovery license entitled for use on cloud-hosted virtual machines.
• A Microsoft Entra ID tenant configured for user sign-in to the virtual machine (VM), with role-based access control (RBAC) defined for the engineering team.
• Network connectivity from users to the Azure region that hosts the VM (for example, internet, ExpressRoute, or site-to-site VPN), with latency suitable for interactive remote graphics.
• A supported operating system on the VM and the AMD Radeon Pro V710 GPU driver appropriate for that OS.
• An Azure virtual network with accelerated networking enabled for low-latency, high-throughput connectivity between users, the VM, and storage.
• Azure Files or Azure NetApp Files storage for CAD models, simulation data, and project artifacts.

Deployment steps

To deploy Ansys Discovery on Azure, complete the following steps:

1. Create an Azure GPU-enabled virtual machine that uses the Standard_NV24ads_V710_v5 SKU. Choose a smaller NVads V710 v5 size if a fractional GPU allocation meets your workload needs.
2. Install the AMD Radeon Pro V710 GPU driver (version Q4-MR1-25.10.17.02) on the VM.
3. Deploy Azure Files storage to host assemblies, metadata, drawing packages, product manufacturing information (PMI), and simulation data.
4. Configure the Azure virtual network with accelerated networking enabled.
5. Create a dedicated Azure virtual machine for the Ansys license server, install the Ansys licensing service on it, and confirm that the GPU VM can reach the license server over the virtual network.
6. Install the Ansys Discovery software on the GPU VM, and configure it to check out licenses from the license server.
7. Install the Ansys Discovery test suites on the GPU VM.

Contributors

Microsoft maintains this article. The following contributors, from Microsoft and Synopsys, originally wrote it:

Principal author:

• Sunita Phanse | Sr. Technical Program Manager

Contributors:

• Madeleine Driver | Lead R&D Engineer at Synopsys
• Roman Walsh | Product Manager at Synopsys

To see nonpublic LinkedIn profiles, sign in to LinkedIn.

Related information

• Azure NVads V710 v5-series
• GPU-accelerated virtual machine sizes
• Install AMD GPU drivers on N-series VMs
• Hub-and-spoke network topology in Azure
• What is Azure NetApp Files?
• Azure Well-Architected Framework