At CES 2025, NVIDIA announced the launch of RTX AI PCs, powered by its new GeForce RTX 50 Series GPUs. These systems enable generative AI models to run locally, transforming digital humans, content creation, and productivity workflows. Built on NVIDIA’s Blackwell architecture, the RTX 50 GPUs deliver up to 3,352 trillion operations per second and introduce FP4 compute, doubling AI inference performance while reducing memory requirements.
The announcement includes NVIDIA NIM microservices, designed to simplify access to AI models for tasks like large language models (LLMs), image generation, and speech recognition. NVIDIA also introduced AI Blueprints, preconfigured workflows for creators and developers, enabling innovative use cases like PDF-to-podcast conversion and 3D-guided image generation.
Jensen Huang, NVIDIA’s founder and CEO, stated that “NIM microservices and AI Blueprints give PC developers and enthusiasts the building blocks to explore the magic of AI,” emphasizing their role in advancing AI technology for everyday use.
Major PC manufacturers, including Acer, Dell, MSI, and others, will launch NIM-ready RTX AI PCs. Starting in February, these systems will support RTX 50 Series GPUs and other NVIDIA hardware, marking a significant step forward in bringing AI-powered tools to desktops.
