NVIDIA Open Sources Run:ai Under Apache 2.0
NVIDIA has open-sourced the complete Run:ai GPU orchestration platform under Apache 2.0, featuring KAI Scheduler and CUDA 13.4.
Stock photo for illustration only, not from the actual event
- NVIDIA releases the entire Run:ai GPU orchestration platform as open source
- Licensed under Apache 2.0 for broader developer access and flexibility
- Includes KAI Scheduler, quota management, and CUDA 13.4 GPU fractioning
- Features a local Kind dev environment for streamlined development
NVIDIA has officially released the full Run:ai GPU orchestration platform as open source under the Apache 2.0 license. This move grants developers and enterprises open access to advanced infrastructure management tools designed to optimize high-performance artificial intelligence workloads on Kubernetes.
The open-source release encompasses several core components and utilities engineered to streamline resource allocation and workload scheduling for modern computing clusters.
- KAI Scheduler for advanced workload scheduling
- CUDA 13.4 supporting GPU fractioning capabilities
- Comprehensive resource quota management systems
- A local Kind dev environment for development and testing
By open-sourcing Run:ai under Apache 2.0, NVIDIA is significantly strengthening the enterprise AI ecosystem on Kubernetes. This transition enables organizations to maximize the utilization of expensive hardware accelerators by efficiently sharing GPU resources across multiple teams and concurrent AI projects.
Stock photo for illustration only, not from the actual event
These tools empower infrastructure administrators to enforce strict usage quotas and fine-tune resource distribution, ultimately reducing the operational complexity of managing large-scale distributed computing environments for machine learning initiatives.
Source: MarkTechPost
Found something wrong in this article? Report an issue with this article
Comments
Leave a Comment