NVIDIA Virtual PC (vPC): Sizing and GPU Selection Guide for Virtualized Workloads

Sizing Guide

• Overview
  
  • What is NVIDIA vPC?
  • About NVIDIA nVector Benchmark
  • NVIDIA vGPU Architecture
• Recommended NVIDIA GPUs for NVIDIA vPC
• Selecting the Right NVIDIA GPU for Virtualization
  
  • NVIDIA GPUs for Virtualization
    
    • NVIDIA RTX PRO 6000 Blackwell Server Edition
      
      • MIG-Backed vGPU Support
    • NVIDIA L40S
    • NVIDIA L40
    • NVIDIA L4
    • NVIDIA A40
    • NVIDIA A16
    • NVIDIA A10
  • Knowledge Worker VDI
• Selecting the Right NVIDIA GPU Virtualization Software
  
  • Summary of Product Features
  • Product Details
    
    • NVIDIA RTX Virtual Workstation
    • NVIDIA Virtual PC
  • Impact of GPU Sharing
    
    • Time-Sliced vGPU Sharing
      
      • Scheduling Options for GPU Sharing
    • MIG-Backed vGPU Sharing
      
      • Performance Allocation
• Sizing Methodology
  
  • vGPU Profiles
    
    • Example Homogeneous vGPU Configurations for NVIDIA A16
    • Example Heterogeneous vGPU Configurations for NVIDIA A16
• Tools
  
  • GPU Profiler
  • NVIDIA System Management Interface (nvidia-smi)
  • VMware esxtop
  • VMware Aria Operations
• Testing Methodology
  
  • Single VM Testing
    
    • Test Environment
    • Test Metrics - Framebuffer Usage
  • Scalability Testing
    
    • Server Utilization Metrics
    • User Experience Metrics
      
      • Latency Metrics
      • Remoted Frames Metrics
      • Image Quality
• Test Findings
  
  • Single VM Multi-Monitor Resolution Test Results
    
    • High Definition (1920x1080) Displays
    • Quad High Definition (2560x1440) Displays
    • 4K (4096x2160) Displays
    • 5K (5120x2880) Display
    • Single VM Multi-Monitor Resolution Summary
  • Multi-VM Scalability Test Results
    
    • Server Utilization Metrics
    • nVector User Experience Metrics
      
      • Frame Rate
      • Latency Metrics
      • Encoder Utilization
      • Image Quality
    • Multi-Monitor Resolution Scalability Summary
• Deployment Best Practices
  
  • Run a Proof of Concept
  • Leverage Management and Monitoring Tools
  • Understand Your Users
  • Use Benchmark Testing
  • Understanding the GPU Scheduler
  • Understanding GPU Channels
    
    • Monitoring GPU Channel Usage During POC
• Summary

Appendix

• Framebuffer Utilization
• Additional Resources
  
  • NVIDIA vPC Resources
  • NVIDIA RTX Virtual Workstation Resources
  • Other Resources
• Support and Services