Accelerating Vision AI on Intel® Core™ Ultra with OpenVINO™ and Ultralytics YOLO

Want to optimise computer vision performance on Intel® Core™ Ultra?

In this session, Intel and Ultralytics show how to deploy Ultralytics YOLO models with OpenVINO, with a real benchmark improvement from 48 FPS to 869 FPS.

In this demo, you’ll see how to:

• Convert Ultralytics YOLO models to OpenVINO format

• Run baseline inference and benchmark across CPU, GPU, and NPU

• Identify bottlenecks in your pipeline

• Optimise throughput and latency for edge deployment

Led by Adrian Boguszewski (Intel) and Francesco Mattioli (Ultralytics), the session walks through a full workflow from model export to performance optimisation, with real results from live testing.

Relevant for developers working in computer vision, edge AI, and performance optimisation across industrial and real-time applications.