KEDA

2 articles tagged with this topic

KServe Breaks Inference Autoscaling Into Three Paths as GPU Costs Face Scrutiny

KServe routes request-, resource-, and event-driven scaling to KPA, HPA, and KEDA, bringing GPU utilization and inference cost into focus.

3d ago2 min read

SageMaker HyperPodKarpenter

Best practices to run inference on Amazon SageMaker HyperPod

AWS details H yperPod inference deployment patterns, claiming up to 40% total cost of ownership reduction for GPU work loads.

Apr 152 min read