NVIDIA Introduces AI Innovation with RTX AI PCs
At CES, NVIDIA unveiled groundbreaking foundation models designed to run locally on RTX AI PCs, powered by the advanced GeForce RTX™ 50 Series GPUs. These models aim to revolutionize digital humans, content creation, and productivity workflows.
The GeForce RTX 50 Series GPUs, built on NVIDIA’s cutting-edge Blackwell architecture, deliver an astounding 3,352 trillion operations per second of AI performance. With 32GB of VRAM and support for FP4 compute, these GPUs double AI inference performance and allow generative AI models to operate within a smaller memory footprint compared to previous generations.
A Legacy in AI Development
GeForce™ GPUs have been a cornerstone of AI innovation. From training the first GPU-accelerated deep learning network, AlexNet, on the GeForce GTX™ 580 in 2012, to being cited in over 30% of AI research papers in 2024, NVIDIA continues to drive the AI revolution. Now, with RTX AI PCs, even enthusiasts can easily explore AI through low-code and no-code tools like AnythingLLM, Langflow, and LM Studio.
NIM Microservices: Simplifying Generative AI Deployment
NVIDIA NIM™ microservices are designed to make deploying generative AI models effortless. These microservices, paired with intuitive graphical user interfaces, provide preconfigured workflows for tasks like creating digital humans and generating content. RTX AI PCs from leading manufacturers such as Dell, ASUS, and Lenovo will be equipped to support these microservices.
Agentic AI Takes Center Stage
NVIDIA also introduced Project R2X, an AI-powered PC avatar capable of assisting with tasks such as information retrieval, document summarization, and video conferencing. This avatar is powered by NVIDIA RTX Neural Faces and the Audio2Face™-3D model, ensuring lifelike animations and enhanced interaction capabilities.
Additionally, NVIDIA announced the Llama Nemotron family of open models, optimized for tasks such as chat, coding, and instruction following. These models, available as NIM microservices, further expand the capabilities of agentic AI.
AI Blueprints: Empowering Developers
To empower developers, NVIDIA introduced AI Blueprints — prebuilt workflows that run locally on RTX PCs. For example, a PDF-to-podcast blueprint enables users to extract content from PDFs and convert it into a podcast. Another blueprint allows creators to use 3D scenes in tools like Blender to guide AI-generated imagery, offering greater artistic control.
These blueprints leverage NIM microservices such as NVIDIA Riva for text-to-speech and Mistral-Nemo-12B-Instruct for language processing, providing seamless integration for developers and enthusiasts alike.
Pioneering Local AI Development
NVIDIA’s innovations are set to redefine local AI development. With support for Windows Subsystem for Linux (WSL), developers can quickly set up and run NIM microservices on Windows 11 PCs. These services integrate effortlessly with popular AI frameworks, ensuring compatibility and ease of deployment across platforms.
For those interested in transformative AI infrastructure, explore how smarter AI launchpads are shaping future innovation. NVIDIA’s advancements align closely with these industry trends, further solidifying their leadership in AI development.
The Future of AI Innovation
Starting February, NIM microservices and AI Blueprints will be available for RTX 50 Series GPUs and other supported hardware, with more additions expected in the future. By enabling local AI capabilities, NVIDIA is not only advancing AI technology but also making it accessible to a broader audience.
Stay tuned for more updates as NVIDIA continues to push the boundaries of AI innovation, empowering developers, creators, and enthusiasts to explore the limitless possibilities of AI.
