Acceleration Made Easy

Accelerated Applications for Adaptive Platforms

Easy Installation

Apps are packaged in Linux deployment frameworks for easy installation over the air via open-source mechanisms

Free Evaluation

Free evaluation with supported documentation, getting started flows, and additional tools and resources –no hardware expertise needed

Develop and Deploy

Begin development at an elevated starting point and deploy to either the starter kit or the production SOM

Easy Updates

Dynamically upgrade to the latest version of the accelerated app, either to the starter kit or the deployed production SOM

How to Explore

1. Select App
Choose an app and explore functionality, specifications, and documentation

2. Download App
Download App for free and evaluate on the KV260, KR260, or KD240 Starter Kit

3. Follow Step-by-Step
Use our Getting Started Guide and be up and running in less than an hour

4. Purchase App
For production deployment of K24 or K26 production SOM (all AMD apps are free)

Kria Accelerated Apps

 Enabling developers to program and differentiate their designs at the software level, without requiring FPGA programming experience.

Vision & Robotics Apps for Kria K26

All Applications
cargo ship

LED Active Markers Tracking

This plug-and-play LED Active Markers Tracking application showcases PROPHESEE’s event-based Metavision technologies. It excels in high-speed detection and tracking of LED light pulses, complete background removal, and real-time 3D pose information, unlocking new ultra-high-speed tracking applications.
View App
LogicTronix object detection

Object Detection and Tracking with Event Vision-Based Sensors

Event Vision-Based (EVB) sensing offers accelerated sensing speeds, improved operation in unpredictable lighting conditions, and decreased communication demands compared to frame-based sensors.1 This application showcases an integrated EVB sensor via MIPI to the AMD Kria™ KV260 Vision AI Starter Kit, executing object detection and tracking for streamlined, end-to-end pipelined acceleration.

View App
Machine Vision
Smart Camera Accelerated Application

Smart Camera

This UltraHD smart camera implements face detection with network and display functionality. It comes with built-in machine learning for applications such as pedestrian detection, face detection, and people counting with local display and RTSP streaming.
View Apps
10GigE Vision Camera Accelerated Application

10GigE Vision Camera

This app implements an industrial camera design using a Sony IMX SLVS-EC sensor and the widely used GigE Vision protocol. GigE Vision enables the user to configure the camera over the network and stream video data, currently measured up to 10Gb/s. Streaming is handled fully in hardware to get maximum data rate and to free the processing unit of any additional image processing tasks.
View App
Machine Vision
pedestrian detection

AI Box with ReID

The AI Box with ReID accelerated application performs distributed, scalable, multi-stream tracking and Re-Identification. The app leverages machine learning for pedestrian tracking and decoding multiple camera streams and performs pedestrian detection and tracking across camera feeds. Common applications include smart cities, retail analytics, and video analytics.
View App
Machine Vision
produce defect detection

Defect Detection

The Defect Detection accelerated application is a machine vision app that automates detection of defects, (e.g., fruits, PCBs), and sorting in high-speed factory pipelines by using Vitis Vision library functions.
View App
Machine Vision
Aupera Face Recognition for Kria SOM Accelerated Application

Aupera Face Recognition

Aupera Facial Recognition Solution is an end-to-end, commercially deployable solution for facial recognition in the field. Equipped with the Aupera proprietary best-in-class trained AI model, the solution has been in field deployment by Tier-1 customers. It comes with built-in machine learning for applications such as face detection, face recognition, mask detection, face recognition with mask, RTSP/RTMP streaming, and ONVIF interfacing.
View App
Machine Vision
ROS 2 multi node

ROS 2 Multi-Node Communication via TSN

Synchronized Real Time Clocks are a key enabler for automation of complex processes and a deterministic behavior of a system with multiple sensors, actuators, and controllers. The Time-Sensitive Networking (TSN) Subsystem from AMD offers Time synchronization and the time-aware transmission of Ethernet Frames with low jitter. Because it comes with two external interfaces, it can be used for larger networks without needing an external TSN switch.
View App
Robotics
robot arm

ROS 2 Perception Node

The ROS 2 Perception Node accelerated application implements a subset of image_pipeline, which is one of the most popular packages in the ROS 2 ecosystem and a core piece of the ROS perception stack. It creates a simple computational graph consisting of two hardware accelerated nodes, resize & rectify, and leverages KRS framework for tracing and benchmarking.
View App

Demonstration Apps

Demonstration apps are similar to accelerated apps except they are intended for evaluation purposes only. Evaluate capabilities on a starter kit and inquire about next steps with the accelerated app provider.

Minerva Systems screen image

Minerva Systems Architect

Minerva Systems Architect helps identify and solve memory-related interference problems during the integration phases of complex, mixed-criticality workloads. The Architect tool provides a framework designed to profile, visualize, and analyze memory accesses performed by user applications.
View App
Machine Vision
AMD Kria DFX

DFX Accelerators

Dynamic Function eXchange (DFX) utilizes flexibility of Programmable Logic (PL) devices, allowing the runtime modification of an operating hardware design. A partitioned design allows one part of the PL to be reconfigured while another part of the system remains running. The DFX Accelerated App illustrates the premise behind DFX by leveraging the ability to time multiplex hardware dynamically on Kria™.
View App
Machine Vision
Robotics
Avnet SMS diagram

Smart Model Select

The Smart Model Select application is a tutorial example on how to create a custom application for Kria™ SOM. The Smart Model Select application features one of the Vitis™ Video Analytics SDK (VVAS) examples, which has been ported to the Kria KV260 Vision AI Starter Kit.
View App
Machine Vision
speech recognition

NLP-SmartVision

The natural language processing (NLP) SmartVision application continuously detects keywords spoken by the user and showcases keyword-based dynamic switching between multiple vision tasks and/or changes the display properties.
View App
Machine Vision
covid prediction

Low-Code Smart Healthcare Platform

The Spline.ai Low-Code Smart Healthcare Platform presented here is a demonstration using pneumonia and COVID-19 deep learning applications. The model is compiled and optimized using the Vitis™ AI software platform to run inference on the Kria™ KV260 starter kit with the Ubuntu 22.04 operating system. This low-code framework is designed to develop applications either as standalone or with a large fleet of Kria K26 SOM-based edge appliances in an AWS IoT Greengrass v2 platform.
View App
Machine Vision
GL Studio High Fidelity HMI Quad Demonstration Application

High Fidelity HMI Quad

The GL Studio HMI Quad Demonstration presents the flexibility and power of GL Studio by highlighting four of the infinite HMI possibilities that can be created using the tool. Each full HD demo runs at 60hz and is packed with interactive widgets that are driven by simulated data. Preview demos one at a time by zooming in, or view all four at once. Medical, Industrial, and Automotive domains are featured. Let the power of GL Studio software and AMD hardware carry your project from prototype to production in record time with best in class runtime performance.
View App
Machine Vision

Drives Apps for Kria K24

All Applications
ROS 2 multi node

ROS 2 Multi-Node Communication via TSN Accelerated Application for KD240

Synchronized Real Time Clocks are a key enabler for automation of complex processes and a deterministic behavior of a system with multiple sensors, actuators, and controllers. The Time-Sensitive Networking (TSN) Subsystem from AMD offers Time synchronization and the time-aware transmission of Ethernet Frames with low jitter. Because it comes with two external interfaces, it can be used for larger networks without needing an external TSN switch.
View App
Vision AI DPU-PYNQ diagram

Vision AI DPU-PYNQ

PYNQ is an open-source productivity framework built with Python, Jupyter, and an extensive ecosystem of associated libraries. It increases the productivity of software and hardware engineers by using the Zynq™ family of devices to build more capable and intelligent systems. The DPU-PYNQ Accelerated Application includes a Vitis™ AI DPU (Deep Learning Processor Unit) with AI inference notebooks ready to run out of the box.
View App
field oriented motor

Adaptable Field-Oriented Control

This high-performance electric drives app brings the right level of integration and density to achieve hard real-time performance into mission-critical motor control applications. By offloading CPU tasks for independent space vector modulation, analog acquisition, and other related motor I/O tasks, this implementation of a rich Field-Oriented Control (FOC) and Sliding Mode SFOC algorithm in HDL offers an extremely versatile platform for learning and designing.
View App
control arm

Field-Oriented Control

Sensor-based control is a key component of many motor system implementations. In the context of brushless DC motors, field-oriented control (FOC) stands as a significant control methodology. The Field-Oriented Control with Position Sensor Accelerated Application from AMD utilizes the Vitis™ Motor Control Library to offer deterministic and low-latency control of synchronous motors. The system integrates quadrature encoder, DC voltage, and current readings to ensure precise speed and torque control over the target motors.
View App
Add Alt Text

Interested in becoming a technology partner?

The Kria App Store offers a powerful platform to market your unique edge applications, algorithms, and IP cores using a standard infrastructure – Docker containers hosted on Docker Hub. The accelerated application format for delivering designs is the easiest way for AMD customers to evaluate your solution and is also streamlined for the application-focused solution developer compared to the traditional IP product development process.

Become a Technology Partner
curved transparent to black top gradient divider
Add Alt Text

Kria System-on-Modules (SOMs)

Kria SOMs, a portfolio designed for edge deployment from starter kit to production, simplify system development, helping you get your product to market faster.

Explore Kria SOMs
Kria System-on-Modules
Footnotes
  1. All performance benefits and/or time savings claims are provided by PROPHESEE and/or LogicTronix Technologies and have not been independently tested or verified by AMD. Performance benefits and time savings are impacted by a variety of variables. Results herein are specific to PROPHESEE and/or LogicTronix Technologies and may not be typical. GD-181