Video by CNCF [Cloud Native Computing Foundation] via YouTube
Can Apple Silicon become a production GPU node for LLM inference on Kubernetes?
At KubeCon + CloudNativeCon North America, Chris Maher of Ashley Furniture Industries will share how his team extended Kubernetes to schedule inference across heterogeneous GPU hardware using pluggable runtime backends.
Learn about model lifecycle management, multi-GPU sharding, memory validation, Prometheus metrics, and the operational tradeoffs between NVIDIA CUDA and Apple Silicon Metal.
Explore the session: https://bit.ly/4jZDJSu
Join us November 9-12 in Salt Lake City.
Register: https://bit.ly/3BDI5XL