Memory-Enforced GPU Sharing in Kubernetes with NVIDIA CUDA 13.4
Source
developer.nvidia.com
Author
Elizabeth Goodman
Date
Why it matters
CUDA 13.4 can now enforce per-workload GPU memory limits through MPS. With the GPU Fractioning Operator and KAI Scheduler, several inferenceRunning a trained model to get answers — the phase where AI is actually used, as opposed to trained.Full definition → workloads can share one GPU on Kubernetes without one exhausting its memory.
Terms in this piece · Glossary
inference — Running a trained model to get answers — the phase where AI is actually used, as opposed to trained.
articleHow NVIDIA OpenShell Reference Architecture Secures and Governs AI Agent Workloads at Scale
articleCode, Test, and Deploy on One Machine with NVIDIA RTX Spark
articleHow to Rewrite Apache Spark UDFs for GPU Acceleration with AI Agents
articleLower the Cost of Building and Running Visual AI Agents with NVIDIA VSS Blueprint 3.3