NVIDIA Unveils NIM Microservices and AI Blueprints for Local AI Development

LAS VEGAS, Nevada — NVIDIA has announced the launch of NIM microservices and AI Blueprints, enabling developers and enthusiasts to build AI agents and creative workflows directly on PCs. The announcement, made at CES 2025, highlights NVIDIA’s latest advancements in generative AI and local AI processing, powered by the new GeForce RTX 50 Series GPUs.

The NIM microservices, optimized for NVIDIA’s Blackwell architecture, allow foundation models to run locally on PCs. These models, which include large language models (LLMs), vision language models, and image generation tools, are accelerated by the RTX 50 Series GPUs, offering up to 3,352 trillion operations per second of AI performance. The GPUs also introduce FP4 compute, doubling AI inference performance and reducing memory requirements for generative AI models.

Jensen Huang, founder and CEO of NVIDIA, stated,