Profiling AI models using Qualcomm AI Engine Direct @QualcommDeveloper
Profiling AI models using Qualcomm AI Engine Direct  @QualcommDeveloper
Uploaded October 2025 | Updated November 2026, 8 hours ago
Unlock the full potential of your AI models on Qualcomm® hardware!

This video walks developers through profiling AI models using the Qualcomm® AI Engine Direct SDK, covering essential steps and best practices for optimising performance on Qualcomm® Kryo™ CPU, Adreno™ GPU, and Hexagon™ NPU.

Chapters include:

Prerequisites before profiling
Qualcomm AI Engine Direct SDK workflow
qnn-net-run
Qualcomm Neural Network linting profiling
Model optimisation example using linting profiling
Identifying the bottleneck
Additional guidelines

Learn how to target specific accelerators, delegate workloads from TensorFlow Lite or ONNX runtime, and use unified APIs to get closer to the silicon for maximum efficiency.

Find out more about the SDK: qualcomm.com/developer/software/qualcomm-ai-engine-direct-sdk

If you have any questions or need further assistance, feel free to leave a comment below. Subscribe: More Qualcomm Developer videos: tinyurl.com/2p8xmcw6

Join our community of developers on Discord to stay up-to-date, find expert support, and access exclusive virtual events: discord.gg/THUPBtskgs
Profiling AI models using Qualcomm AI Engine DirectUnlocking the potential of Edge AI: Innovations and Best Practices I Qualcomm at AWS:reinvent 2024Explore Qualcomm Software CenterQualcomm® Linux® sample applications: learn to stream & display multiple camera feedsDeveloper Tech Talks: Introducing Qualcomm RB3 Gen 2 PlatformQualcomm AI Inference Suite: IntroductionMeet and Greet: MIT Reality Hack 2025 winnersLive Build Along: Using TensorFlow to accelerate models on Qualcomm IoT devicesAccenture x Qualcomm edge AI inference demo at AWS re:invent 2024Run AI where the data lives📍Cephable x Qualcomm AI hub – empowering users through gesture control | Microsoft Build DemoIntroduction to the FoundriesFactory™ Platform
Qualcomm Developer |

Profiling AI models using Qualcomm AI Engine Direct

SHARE TO X SHARE TO REDDIT SHARE TO FACEBOOK WALLPAPER