NVIDIA NIM Microservices and AI Blueprints help developers and enthusiasts build AI agents and creative workflows on the PC
CES—NVIDIA today announced a foundational model running locally on NVIDIA RTX™ AI PCs that powers digital humans, content creation, productivity, and development.
These models are delivered as NVIDIA NIM™ microservices and are accelerated by the new GeForce RTX™ 50 Series GPUs with up to 3.352 trillion operations per second of AI performance and 32 GB of VRAM. Built on the NVIDIA Blackwell architecture, the RTX 50 series is the first consumer GPU to add support for FP4 compute, delivering 2x faster AI inference performance compared to previous generation hardware and faster generation of AI models. can be executed locally with a smaller memory footprint.
GeForce™ has long been the essential platform for AI developers. The first GPU-accelerated deep learning network, AlexNet, was trained on a GeForce GTX™ 580 in 2012. And last year, more than 30% of published AI research papers mentioned the use of GeForce RTX.
Now, with Generative AI and RTX AI PCs, anyone can become a developer. A new wave of low-code and no-code tools such as AnythingLLM, ComfyUI, Langflow, and LM Studio allow enthusiasts to use AI models in complex workflows through a simple graphical user interface.
NIM microservices connected to these GUIs make it easy to access and deploy modern generative AI models. Built on NIM microservices, NVIDIA AI Blueprints provide easy-to-use, preconfigured reference workflows for digital humans, content creation, and more.
To meet the growing demand from AI developers and enthusiasts, all top PC manufacturers and system builders are launching NIM-enabled RTX AI PCs powered by GeForce RTX 50 Series GPUs.
“AI is progressing at the speed of light, from perceptual AI to generative AI and now agent AI,” said Jensen Huang, Founder and CEO of NVIDIA. “NIM Microservices and AI Blueprints provide PC developers and enthusiasts with the building blocks to explore the magic of AI.”
Make AI easy
The underlying model, a neural network trained on vast amounts of raw data, is the building block of generative AI.
NVIDIA releases a pipeline of NIM microservices for RTX AI PCs from top model developers including Black Forest Labs, Meta, Mistral, and Stability AI. Use cases span large-scale language models (LLM), vision language models, image generation, speech, embedded models for search augmented generation (RAG), PDF extraction, and computer vision.
“GeForce RTX 50 Series GPUs with FP4 compute unlock a wide range of models that can run on PCs that were previously limited to large data centers,” said Robin Rombach, CEO of Black Forest Labs. says Mr. “By making FLUX an NVIDIA NIM microservice, we can deploy AI faster and reach more users, while delivering incredible performance.”
NVIDIA today also announced the Llama Nemotron family of open models that deliver high accuracy across a wide range of agent tasks. The Llama Nemotron Nano model is delivered as a NIM microservice for RTX AI PCs and workstations and excels at agent AI tasks such as instruction following, function calls, chat, coding, and math.
NIM microservices contain the key components for running AI on PCs and are optimized for deployment across NVIDIA GPUs, including RTX PCs, workstations, and the cloud.
Developers and hobbyists can now quickly download, set up, and run these NIM microservices on Windows 11 PCs with Windows Subsystem for Linux (WSL).
“AI is rapidly driving innovation on Windows 11 PCs, and the Windows Subsystem for Linux (WSL), along with Windows Copilot Runtime, provides a great cross-platform environment for AI development on Windows 11.” said Pavan Davuluri, corporate vice president of Windows. Microsoft. “NVIDIA NIM microservices optimized for Windows PCs enable developers and enthusiasts to quickly integrate AI models into their Windows apps, further accelerating the deployment of AI capabilities to Windows users.”
NIM microservices running on RTX AI PCs are compatible with leading AI development and agent frameworks such as AI Toolkit for VSCode, AnythingLLM, ComfyUI, CrewAI, Flowise AI, LangChain, Langflow, and LM Studio. Developers can connect applications and workflows built on these frameworks to AI models running NIM microservices through industry-standard endpoints, so they can be used across clouds, data centers, workstations, and PCs. You’ll be able to use the latest technology with a unified interface.
Enthusiasts can also experience various NIM microservices using the upcoming release of the NVIDIA ChatRTX technology demo.
Putting a face on Agentic AI
To demonstrate how hobbyists and developers can use NIM to build AI agents and assistants, NVIDIA today previewed Project R2X. It’s a vision-enabled PC avatar that can put information at your fingertips, assist with desktop apps and video conferencing, and read and summarize documents. , etc.
Avatars are rendered using NVIDIA RTX Neural Faces, a new generative AI algorithm that enhances traditional rasterization with fully generated pixels. The face is then animated by a new diffusion-based NVIDIA Audio2Face™-3D model that improves lip and tongue movements. R2X can connect to cloud AI services such as OpenAI’s GPT4o and xAI’s Grok, NIM microservices and AI blueprints such as PDF Retriever and alternative LLMs through developer frameworks such as CrewAI, Flowise AI, and Langflow. . Sign up for Project R2X updates.
AI Blueprint comes to PC
NIM microservices are also available to PC users through AI Blueprints (see AI workflows that can run locally on RTX PCs). These blueprints allow developers to create podcasts from PDF documents or generate stunning images based on 3D scenes.
The PDF to Podcast Blueprint extracts text, images, and tables from PDFs to create podcast scripts that users can edit. You can also generate complete audio recordings from your script using the audio available in your blueprint or based on your user’s audio samples. Additionally, users can have real-time conversations with AI podcast hosts to learn more about specific topics.
This blueprint uses NIM microservices such as Mistral-Nemo-12B-Instruct for language, NVIDIA Riva for text-to-speech and automatic speech recognition, and the NeMo Retriever collection of microservices for PDF extraction.
3D Guided Generation AI’s AI Blueprints give artists more control over image generation. AI can generate great images from simple text prompts, but controlling the composition of an image using just words can be difficult. This Blueprint allows creators to guide AI image generation using simple 3D objects laid out in a 3D renderer such as Blender. Artists can create 3D assets manually or generate them using AI, place them in the scene, and set the 3D viewport camera. A prepackaged workflow powered by FLUX NIM microservices then uses the current configuration to generate high-quality images that match the 3D scene.
NVIDIA NIM Microservices and AI Blueprints will be available starting in February with initial hardware support for GeForce RTX 50 Series, GeForce RTX 4090 and 4080, and NVIDIA RTX 6000 and 5000 Professional GPUs. Additional GPUs will be supported in the future.
NIM-enabled RTX AI PCs are available from Acer, ASUS, Dell, GIGABYTE, HP, Lenovo, MSI, Razer, Samsung, and local system builders Corsair, Falcon Northwest, LDLC, Maingear, Mifcon, Origin PC, PCS, Scan It will be available from. .
Join NVIDIA at CES to learn more about how NIM microservices, AI Blueprints, and NIM-enabled RTX AI PCs are accelerating generative AI.
