在 Kubernetes 中透過 NVIDIA CUDA 13.4 實現記憶體限制的 GPU 共享
Memory-Enforced GPU Sharing in Kubernetes with NVIDIA CUDA 13.4
NVIDIA CUDA 13.4 支援在 Kubernetes 中以記憶體限制方式共享 GPU。許多 AI 工作負載只需使用部分 GPU；與每個工作負載獨佔整張 GPU 相比，共享 GPU 可讓平台團隊利用現有基礎設施服務更多工作負載。
Many AI workloads need only a fraction of a GPU. Giving each one an entire device can leave capacity idle; sharing GPUs lets platform teams serve more workloads...
Many AI workloads need only a fraction of a GPU. Giving each one an entire device can leave capacity idle; sharing GPUs lets platform teams serve more workloads with the infrastructure they already have. NVIDIA GPUs already support several approaches to sharing. Multi-instance GPU (MIG) splits a GPU into hardware-isolated instances that are transparent to the application…
來源：NVIDIA Technical Blog：Agentic AI / Generative AI · developer.nvidia.com