Overview

The Qualcomm Innovators Development Kit (QIDK) is designed to empower developers to build, test, and scale applications on Snapdragon® mobile platforms—ranging from on device AI workloads to low level system exploration. QIDK combines powerful hardware, comprehensive software tools, and developer support, enabling students, academic researchers, and developers to prototype a wide range of applications. These include on device AI solutions, system level software such as device drivers, and hands‑on exploration of the internal architecture of modern Snapdragon processors that power today’s flagship smartphones. Built on premium Snapdragon system on chip (SoC) platforms, QIDK provides access to the same features, capabilities, and performance characteristics relied upon by leading device manufacturers—now made available to the broader developer community.

../_images/QIDK_Kit.jpg

Key Highlights

  • Rapid transition from idea to prototype across Snapdragon powered devices

  • Edge AI development on Snapdragon platforms with CPU, GPU, and NPU support

  • Hardware Expansion Cards for real-world prototyping and diverse use cases.

    • Camera Card – Advanced imaging & video capabilities; rear-facing sensor

    • Sensor Card – Pressure, Hall, Proximity, Magnetometer, Accelerometer, Gyroscope

    • Display Card – FHD+ LCD Capacitive Touch

    • Audio & Controls – Extended audio connectors; volume & power buttons

  • On device deployment of modern generative AI models, including LLMs and diffusion based image generation models

  • Easy onboarding of pre optimized models from Qualcomm AI Hub

  • End-to-End AI Solutions Available on QIDK GitHub.

    • Vision: Object detection, image enhancement, super resolution, pose estimation

    • NLP: Sentiment analysis, question answering, automatic speech recognition

    • Generative AI: LLM-based chat applications and diffusion-based image generation

Supported Snapdragon Platforms

QIDK is built on premium Snapdragon® mobile platforms and supports development and deployment across the Snapdragon 8 series, including recent and current generations such as:

  • Snapdragon® 8 Elite Gen 5

  • Snapdragon® 8 Elite Gen 4

  • Snapdragon® 8 Gen 3

  • Snapdragon® 8 Gen 2

These platforms provide access to Snapdragon’s heterogeneous compute architecture, including CPU, GPU, and NPU, enabling efficient on-device AI execution.

Supported Operating System

QIDK runs only on Android operating system, providing a application development environment for building, testing, and deploying on-device AI applications.

Learning Resources

Video Tutorials

Hardware Development Kit Unboxing

Unboxing and overview of the QIDK hardware kit.

https://www.youtube.com/watch?v=59kN40__apo
AI on QIDK

Structured course with hands-on labs on Qualcomm Academy.

https://academy.qualcomm.com/course-catalog/AI-on-QIDK
AI Upskilling Certificate

Technical Foundations certificate program on Qualcomm Academy.

https://academy.qualcomm.com/course-catalog/AI-Upskilling-Certificate-Technical-Foundations

Documentation

QAIRT Documentation

Qualcomm AI Runtime (QAIRT) reference documentation.

https://docs.qualcomm.com/doc/80-63442-10/topic/general_overview.html
AIMET Documentation

AI Model Efficiency Toolkit (AIMET) documentation and overview.

https://quic.github.io/aimet-pages/releases/latest/overview/index.html

Open-Source Solutions

Qualcomm Sample AI/ML E2E Solutions

Open-source Vision, NLP, and GenAI E2E apps on GitHub.

https://github.com/qualcomm/qidk/tree/master/Solutions
Qualcomm Sample GenAI E2E Solutions

Generative AI solutions including LLMs and diffusion models on GitHub.

https://github.com/qualcomm/qidk/tree/master/GenAI-Solutions

Support

For bug reports and feature requests, raise an issue on our GitHub Issues page.

For questions or suggestions, reach out to us at qidk@qti.qualcomm.com.

Community

Built something with QIDK? Share your solutions on GitHub and tag them with #QIDK so the community can discover and learn from your work.

Share your experience on LinkedIn using #QIDK to connect with fellow developers and the Qualcomm team.

Community Projects