设计中心
培训
Unlocking AI/ML Performance on NXP’s i.MX 93 Applications Processor

Unlocking AI/ML Performance on NXP’s i.MX 93 Applications Processor

About This Training

Every bit of performance counts when developing machine learning applications for the edge. NXP i.MX 93 applications processor, featuring the Arm® Ethos™-U65 microNPU, delivers fast, energy-efficient, ML inference processing. With NXP’s eIQ® toolkit, you can seamlessly convert, optimize, and deploy your models on the NPU with ease.

In this training, we’ll dive into the key considerations for maximizing AI/ML performance on the i.MX 93 MPU. You’ll learn how to avoid common model conversion pitfalls, navigate trade-offs in model parameters, quantization, and ML frameworks, and apply advanced techniques to further enhance performance.

Presentation

What You Will Learn

Key features of the i.MX 93 processor and how the integrated Arm Ethos-U65 microNPU accelerates edge AI workloads
Comparing post-training quantization formats and techniques to improve model efficiency
Using the Vela compiler in the eIQ environment to optimize models for the NPU
Enhancing performance through advanced optimizations, including camera pipeline tuning and distributing operations across processor components
Real-world applications demonstrating these strategies in action