AI Model Distribution Challenges and Best Practices
Talkat KubeCon + CloudNativeCon China 2025 · on Tuesday, 2025-06-10· at 11:00· Level 19 | Crystal Court I·
summary
Practical approaches for distributing large AI/ML models in cloud-native environments, focusing on Kubernetes-native workflows, model storage/transfer optimization, and cross-cluster/region consistency. Covers handling very large models (hundreds of GBs to TBs), challenges of distributed inference and prefill-decode architectures, reinforcement-learning post-training updates, and the use of standards like OCI artifacts or specialized registries to manage versioned model delivery.
AI-written summary · about
related talks
37mHow to Accelerate Model Training and Eliminate the I/O bottleneck for the Cloud
43mProblems and Solutions with AI Model Deployments in k8s
36mMulti-Cloud Machine Learning Data and Workflow with Kubernetes
15mAI Model Inference on K8s
35mStore AI/ML models efficiently with OCI Artifacts
28mProductionizing Machine Learning Models at Scale with Kubernetes
29mAI/MLOps for Busy People: A Field Guide To Implementing Cloud Native AI/ML
32mDesign Patterns for Running AIML and Bigdata Workloads on Kubernetes