NVIDIA has announced DSX OS, an open-source, modular software layer for its DSX platform aimed at operating AI factories at gigawatt scale. Key components include DSX Exchange (MQTT-based IT/OT communication hub with MCP server support for agentic interfaces), MaxLPS and DSX Flex for power and efficiency optimization (enabling up to 40% more GPUs within a fixed power budget), NVIDIA Infra Controller (NICo) for bare-metal lifecycle management with BlueField DPU-enforced tenant isolation, AICR for version-locked runtime configurations, NVSentinel for Kubernetes-native GPU fault detection and automated remediation, Fleet Intelligence for fleet-wide visibility, and KAI Scheduler, Run:ai, Dynamo, Grove, and NVCF for intelligent workload scheduling and distributed inference. All components are available on GitHub and designed for incremental adoption into existing stacks.

7m read timeFrom developer.nvidia.com
Post cover image
Table of contents
Why DSX OS matters to the AI factory ecosystemHow DSX OS enables gigawatt-scale AI factoriesGetting started
1.2K Impressions