NVIDIA Unveils RTX PRO Blackwell GPUs and NIM Microservices at GTC 2025

Rebeca Moen
Mar 19, 2025 02:09

NVIDIA introduces RTX PRO Blackwell series GPUs and NIM microservices at GTC 2025, enhancing AI development on PCs and workstations with new AI-powered tools.

NVIDIA has announced a new line of RTX PRO Blackwell series GPUs and NVIDIA NIM microservices at the NVIDIA GTC 2025 global AI conference. These innovations are aimed at accelerating AI development on PCs and workstations, offering developers powerful tools to integrate AI into creative projects, applications, and games.

Advanced Hardware for AI-Driven Workflows

The RTX PRO Blackwell series, designed to handle intensive AI-driven workloads, includes both desktop and laptop GPUs, as well as a data center GPU. This lineup aims to support demanding applications like AI agents, simulation, 3D design, and visual effects, delivering enhanced performance and scalability essential for professionals in various industries.

The series includes:

• Desktop GPUs: RTX PRO 6000 Blackwell Workstation Edition, RTX PRO 6000 Blackwell Max-Q Workstation Edition, RTX PRO 5000 Blackwell, RTX PRO 4500 Blackwell, and RTX PRO 4000 Blackwell.
• Laptop GPUs: RTX PRO 5000 Blackwell, RTX PRO 4000 Blackwell, RTX PRO 3000 Blackwell, RTX PRO 2000 Blackwell, RTX PRO 1000 Blackwell, and RTX PRO 500 Blackwell.
• Data center GPU: RTX PRO 6000 Blackwell Server Edition.

NVIDIA NIM Microservices for Enhanced AI Capabilities

NVIDIA also introduced NIM microservices, providing developers with high-performance AI models optimized across NVIDIA GPUs. These microservices enable the creation of AI-powered applications, productivity tools, and advanced content-creation workflows. They incorporate top community and NVIDIA-built models supporting large language models (LLMs), images, speech, and retrieval-augmented generation (RAG).

Boosting AI Development with CUDA-X Libraries

The CUDA-X libraries, part of NVIDIA’s AI-powered toolkit, offer significant performance enhancements for data processing and machine learning tasks. The latest release, cuML 25.02, accelerates scikit-learn, UMAP, and HDBSCAN algorithms with no code changes required, facilitating faster exploratory data analysis and model development.

ChatRTX and AI Blueprints

In addition to hardware and microservices, NVIDIA has updated ChatRTX, a demo app that personalizes large language models to user content, providing contextually relevant answers swiftly and privately. This update integrates support for NVIDIA NIM microservices, expanding its capabilities.

NVIDIA AI Blueprints, announced earlier this year, provide advanced AI reference workflows built on NIM, allowing developers to create diverse AI-powered applications. These resources are set to be available soon on build.nvidia.com, offering source code, sample data, and documentation for customization.

For more detailed announcements and updates from the NVIDIA GTC 2025, visit the official NVIDIA blog.

Image source: Shutterstock