NVIDIA
NVIDIA
nsight-copilot-rag
Container
NVIDIA
NVIDIA
nsight-copilot-rag

CUDA RAG backend for NVIDIA Nsight Copilot offline deployment on DGX Spark

Provides retrieval-augmented generation for Nsight Copilot.

It manages a local vector database (Milvus), ingests CUDA documentation and code context, and serves semantic search results to the Nsight Copilot.

Publisher
NVIDIA
NVIDIA
Latest Taglatest
UpdatedJune 10, 2026 UTC
Compressed Size418.49 MB
Multinode SupportNo
Multi-Arch SupportYes