NVIDIA® has introduced NVIDIA Rubin CPX, a new class of GPU designed to power massive-context AI workloads. With this innovation, AI systems can process million-token software coding projects and generative video tasks with unprecedented speed and efficiency.
The Rubin CPX is engineered to work seamlessly with NVIDIA Vera CPUs and Rubin GPUs inside the newly launched NVIDIA Vera Rubin NVL144 CPX platform. This fully integrated NVIDIA MGX system delivers 8 exaflops of AI compute, offering 7.5x more AI performance than the previous NVIDIA GB300 NVL72 platform. It also provides 100TB of high-speed memory and 1.7 petabytes per second of bandwidth within a single rack. For enterprises already invested in Vera Rubin NVL144, a standalone Rubin CPX compute tray will be available for easy integration.
“The Vera Rubin platform will mark another leap in the frontier of AI computing introducing both the next-generation Rubin GPU and a new category of processors called CPX,” said Jensen Huang, founder and CEO of NVIDIA. “Just as RTX revolutionized graphics and physical AI, Rubin CPX is the first CUDA GPU purpose-built for massive-context AI, where models reason across millions of tokens of knowledge at once.”
Also Read: Acer unveils Compact Veriton GN100 AI Mini Workstation Powered by NVIDIA GB10 Superchip
With Rubin CPX, enterprises can unlock new levels of long-context AI performance and monetization opportunities, transforming AI assistants into advanced systems capable of understanding and optimizing large-scale software projects.
In video applications, Rubin CPX tackles one of the industry’s toughest challenges processing up to one million tokens for a single hour of video. By combining video encoders, decoders, and long-context inference into one chip, it enables breakthroughs in video search, generative content creation, and long-form AI video workflows.
Powered by the NVIDIA Rubin architecture, the GPU leverages a cost-efficient monolithic die packed with advanced NVFP4 computing resources. It is optimized for extreme energy efficiency and high-performance AI inference, making it a critical tool for next-generation workloads.
Key Advancements of Rubin CPX
-
30 petaflops of NVFP4 compute for peak performance and precision.
-
128GB of GDDR7 memory for cost-effective, high-throughput workloads.
-
3x faster attention capabilities compared to NVIDIA GB300 NVL72, enabling smoother long-sequence processing.
-
Multiple configurations, including integration with NVIDIA Quantum-X800 InfiniBand or NVIDIA Spectrum-X™ Ethernet with Spectrum-XGS and ConnectX®-9 SuperNICs™.
-
A business model that offers $5 billion in token revenue for every $100 million invested.
Industry Adoption and Use Cases
Early adopters across industries are exploring how Rubin CPX can transform large-scale AI initiatives.
Cursor, an AI-driven software company, plans to integrate Rubin CPX into its code editor platform.
“With NVIDIA Rubin CPX, Cursor will be able to deliver lightning-fast code generation and developer insights, transforming software creation,” said Michael Truell, CEO of Cursor. “This will unlock new levels of productivity and empower users to ship ideas once out of reach.”
Runway, a leader in generative AI for video, will leverage Rubin CPX to accelerate cinematic content creation.
“Video generation is rapidly advancing toward longer context and more flexible, agent-driven creative workflows,” said Cristóbal Valenzuela, CEO of Runway. “We see Rubin CPX as a major leap in performance, supporting these demanding workloads to build more general, intelligent creative tools. This means creators from independent artists to major studios can gain unprecedented speed, realism and control in their work.”
Magic, a company focused on foundation models for AI agents, is also embracing Rubin CPX. “With a 100-million-token context window, our models can see a codebase, years of interaction history, documentation and libraries in context without fine-tuning,” said Eric Steinberger, CEO of Magic. “This enables users to coach the agent at test time through conversation and access to their environments, bringing us closer to autonomous agentic experiences. Using a GPU like NVIDIA Rubin CPX greatly accelerates our compute workloads.”
