Skip to main content
Ctrl+K
AMD Logo
Vitis AI
  • Community

Vitis AI 5.1 documentation

Release Notes

  • Release Notes

Quick Start

  • System Requirements
  • Software Installations
  • Hardware Readiness
  • Execute Sample Model

Tutorials

  • Docker Samples and Demos
  • Optimizing YOLOx Execution with NPU and PL on VEK280
  • Deploying a Custom ResNet18 on Vitis AI NPU
  • Deploying a Custom YOLOv5s on Vitis AI NPU

Customization Opportunities

  • Bring your Own Model (BYOM)
  • Applications
  • Reference Design
    • Default Reference Design
    • Multi-Instance NPU Support
  • Runtime Processes and APIs
  • VART ML C++ APIs
    • Class vart::BaseRunner
    • Class vart::Logger
    • Class vart::Runner
  • VART ML Python APIs
    • runner
  • VART X APIs Architecture Overview
  • VART X APIs Application Developer Guide
  • VART X APIs
    • Class vart::Device
    • Class vart::InferResult
    • Class vart::Memory
    • Class vart::MetaConvert
    • Class vart::Overlay
    • Class vart::PLKernel
    • Class vart::PostProcess
    • Class vart::PreProcess
    • Class vart::VideoFrame

Performance Analyzer

  • Performance

Troubleshooting

  • Troubleshooting

Additional Information

  • License for NPU IP
  • Supported Models

References

  • References

FAQ

  • FAQ
  • <no title>

  • VART X APIs Architecture Overview
    • VART X API Design
    • VART X API Modules
  • VART X APIs Application Developer Guide
    • Application Overview
    • Initialization
    • Main Processing
    • Preprocessing
    • Running Inference
    • Running PLKernel
    • Post-processing
    • InferResult
    • Visualizing Model Predictions
    • Postprocessing Use Cases
    • Configuration File of x_plus_ml_app
    • Enabling Zero Copy Using x_plus_ml_app
    • Multi-instance Support of x_plus_ml_app
    • Introduction
    • Creating Custom Post Processing using supported InferResult
    • Creating Custom Post Processing using custom InferResult
    • Custom Post Processing with custom InferResult
    • Creating Custom MetaConvert
    • Compiling Custom InferResult Implementation as a Separate Library
    • Compiling Custom PostProcess Implementation as a Separate Library
    • Compiling Custom MetaConvert Implementation as a Separate Library
  • VART X APIs
    • Class vart::Device
    • Class vart::InferResult
    • Class vart::Memory
    • Class vart::MetaConvert
    • Class vart::Overlay
    • Class vart::PLKernel
    • Class vart::PostProcess
    • Class vart::PreProcess
    • Class vart::VideoFrame

previous

File runner.py

next

VART X APIs Architecture Overview

Last updated on October 15, 2025.