← Back
NVIDIA Blog

NVIDIA Expands Open Source Commitment With OpenShell Contribution and New AI Infrastructure

•4 min read•
#agents#deployment#nvidia
Level:Advanced
For:AI Engineers
✦TL;DR

NVIDIA announced the open‑source donation of OpenShell, a runtime for securing and governing autonomous agents, to the CNCF Sandbox, and contributed to the Agent Substrate framework for scaling long‑running agent workloads on Kubernetes. New components of the NVIDIA DSX AI Factory Platform—including NVIDIA Kubernetes Engine, Kubernetes Recipes, and Snapshot—were also unveiled to simplify GPU‑accelerated AI deployment at scale. These releases give engineers deterministic policy controls, extensible integration points, and shared‑pool scheduling for AI agents, accelerating secure, large‑scale agent orchestration on Kubernetes.

⚡ Key Takeaways

  • OpenShell applies deterministic controls to agent activity and has received tens of thousands of contributions in less than a year.
  • Agent Substrate preserves agent environments between tasks and integrates with NVIDIA Dynamo for model inference and KAI Scheduler for shared cluster resource allocation.
  • NVIDIA DSX AI Factory Platform adds NVIDIA Kubernetes Engine, Kubernetes Recipes, and Snapshot to manage and scale GPU infrastructure on Kubernetes.
  • OpenShell’s extensible architecture lets teams connect custom compute, inference providers, credential storage, and monitoring tools.
  • Hundreds of organizations have put OpenShell to use, indicating early adoption but requiring integration with existing enterprise systems.
💡 Why It Matters

Engineers can now enforce fine‑grained security policies on AI agents while leveraging shared‑pool Kubernetes scheduling, reducing idle GPU costs and simplifying large‑scale agent deployments.

✅ Practical Steps

  1. Integrate OpenShell into your Kubernetes workloads to enforce deterministic policy controls on agent tool usage and inference service access.
  2. Deploy Agent Substrate alongside NVIDIA Dynamo and KAI Scheduler to preserve agent state across tasks and enable compute suspension/resumption for idle agents.

Want the full story? Read the original article.

Read on NVIDIA Blog ↗

More like this

Introducing workload recovery and expanded infrastructure checks on Amazon SageMaker HyperPod

AWS ML Blog•#deployment

Introducing Unity Gateway Insights

Databricks Blog•#deployment

Introducing Gemini 4 Neon

Google DeepMind•#llm

DaisyRose-27B: Pushing the Frontier of Post-Training Efficiency

Hugging Face Blog•#llm

EXPLORE AI NEWS

Daily hand-picked stories on LLMs, RAG, agents and production AI — curated for engineers who ship.

BROWSE NEWS

GET THE WEEKLY DIGEST

Join engineers getting the Monday signal-over-noise AI breakdown. No spam, unsubscribe anytime.

LEARN AI ENGINEERING

Curated courses, research papers, repos and tutorials built for engineers leveling up in AI.

START LEARNING