ArcGIS Pro is Esri’s flagship desktop GIS application, built to handle rich mapping, 3D visualization, and advanced spatial analysis. Traditionally, these tasks required high-performance desktop workstations. Increasingly, however, organizations are deploying ArcGIS Pro in virtualized environments—either on-premises in their own data centers or in the cloud.
For GIS professionals, virtualization means you don’t always need a powerful machine at your desk. You can log in from virtually anywhere and still access the full power of ArcGIS Pro. For IT administrators, virtualization centralizes management, ensures data security, and allows resources to scale more efficiently.
This guide introduces the hardware resources ArcGIS Pro needs, explains supported and unsupported virtualization models, and provides an overview of both on-premises and cloud-based virtualization options.
Core Resource Requirements
ArcGIS Pro depends on three key resources to run smoothly. These requirements matter whether you are connecting from a workstation, a virtual desktop in your organization’s data center, or a cloud-hosted machine.
CPU (Processor Power)
– What IT needs to know: ArcGIS Pro is multithreaded, so it scales across multiple CPU cores. Virtual CPUs (vCPUs) should be allocated generously to avoid bottlenecks.
– What GIS users will notice: A faster CPU means quicker analysis, faster geoprocessing, and overall responsiveness when working with large datasets.
RAM (Memory)
– What IT needs to know: Allocate memory in line with Esri’s minimum and recommended specs, plus extra headroom for large projects. Avoid overcommitment in multi-user environments.
– What GIS users will notice: More RAM means smoother handling of large layers, faster switching between maps and scenes, and fewer performance slowdowns.
GPU (Graphics Card)
– What IT needs to know: ArcGIS Pro relies heavily on DirectX and OpenGL for visualization. Two GPU models are supported:
• vGPU (shared GPU): Multiple VMs share one GPU; offers good scalability.
• GPU passthrough: One GPU is dedicated to one VM; offers top performance but reduces scalability.
– What GIS users will notice: GPU power directly affects map rendering, panning, zooming, and especially 3D scenes. With a weak or missing GPU, maps may stutter or lag. With a properly configured vGPU or passthrough GPU, performance is as smooth as a local workstation.
– Beyond graphics: Some ArcGIS Pro geoprocessing tools (for example, tools in the Spatial Analyst extension) can leverage NVIDIA CUDA technology for general-purpose GPU acceleration. This means that certain workflows may run significantly faster when GPUs with CUDA support are available.
For additional information on supported NVIDIA GPUs see the following blog:
Unsupported: Remote Desktop Session Hosts (RDSH)
ArcGIS Pro is not supported on Remote Desktop Session Hosts (RDSH)—a session-based virtualization model where many users share a single Windows Server OS.
– Why IT should care: RDSH lacks vGPU scheduling, causing GPU resources to be consumed unevenly. As more users log in, performance degrades for everyone.
– What GIS users will notice: If your ArcGIS Pro session feels sluggish or crashes when more people connect, you may be on an unsupported RDSH setup. Ask for a proper virtual desktop (VDI) instead.
On-Premises Virtualization
Many organizations deploy ArcGIS Pro in on-premises VDI environments using VMware, Citrix, or Nutanix.
– What IT needs to know: These solutions support NVIDIA vGPU technology, allowing GPU resources to be shared across VMs. Proper sizing of CPU, RAM, and GPU is critical.
– What GIS users will notice: Logging into a VDI feels just like using a powerful workstation, but the “workstation” lives in your organization’s data center. This approach is common where sensitive data must remain internal.
⚠️ Note: ArcGIS Pro is not recommended on Citrix Virtual Apps (published apps). It must run in a full virtual desktop.
Cloud Virtualization
Cloud providers now offer GPU-backed instances that work well with ArcGIS Pro.
Supported Platforms
- Amazon Web Services (AWS)
• Microsoft Azure
• Google Cloud Platform (GCP)
• Oracle Cloud
What IT needs to know
– GPU resources can be assigned as shared or passthrough, just like on-premises.
– Remote display protocols (RDP, Amazon DCV, HTML5, etc.) deliver rendered graphics to end-user devices.
– Customers are responsible for selecting the right VM sizes (CPU, RAM, GPU).
What GIS users will notice
– You can access a full Windows desktop with ArcGIS Pro from nearly any device.
– Performance is workstation-like, provided you have a solid internet connection.
– Cloud desktops are especially useful for remote work, contractors, or seasonal projects that need temporary scaling.
Summary
ArcGIS Pro can run successfully in both on-premises and cloud-based virtualized environments—as long as system resources are allocated properly and unsupported models like RDSH are avoided.
– For IT administrators, this means careful VM sizing, GPU configuration, and infrastructure management.
– For GIS professionals, this means you get the flexibility to work from virtually anywhere, without sacrificing the smooth, high-performance experience you expect from ArcGIS Pro.
Whether your organization chooses an on-premises VDI for security and control or cloud virtualization for scalability and flexibility, properly configured environments ensure that ArcGIS Pro delivers the same power as a high-end desktop workstation.
For additional information on supported Virtualization On-Premises and Cloud Provided options the ArcGIS Pro System Requirements – Virtualization Section:
https://pro.arcgis.com/en/pro-app/latest/get-started/virtualization-overview.htm
Article Discussion: