AI Everywhere– accelerate your development from PC to cloud
summary
Shows concrete plumbing: use ONNX Runtime + DirectML and the OpenVINO Execution Provider to run generative models on PC hardware (CPU/GPU/NPU), with compiled-model caching and examples: a mini 53B-style small language model running on DML/GPU and a latent diffusion image model producing 512×512 images in 1–4 steps via OpenVINO. Shows fleet deployment: GitOps pushes OpenVINO model servers to ARC-enabled Kubernetes at Contoso stores for people-counting and checkout-scaling. Cloud side demos Engine X as a service (NGINX Plus) on Azure with 200+ metrics in Azure Monitor and upcoming F5 WAF integration.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 1:01AI acceleration on Xeon CPUs
- 3:49Engine X on Azure
- 8:38Adaptive cloud approach
- 11:10AI use cases in retail
- 18:59Edge AI with ONNX Runtime
- 27:01Security and scalability in AI
- 34:27Q&A
related talks
36mApplied AI on the edge
30mScalable ML acceleration with ONNX Runtime
20mComputer Vision at the Edge with OpenVINO
30mCloud? No Thanks! I’m Gonna Run GenAI on My AI PC
05mONNX Runtime
23mFoundry Local: Cutting-Edge AI experiences on device with ONNX Runtime/Olive
23mOptimize your network inference time with OpenVINO
33mMachine Learning in the Browser: Fast Iteration with ONNX & WebAssembly