On September 9, 2025, NVIDIA revealed its new Rubin CPX GPU, a purpose-built accelerator designed to handle long-context AI tasks like million-token coding and generative video. It’s part of the Vera Rubin NVL144 CPX platform, offering 8 exaflops of AI compute, 100 TB fast memory and 1.7 PB/s bandwidth—and aimed at maximizing efficiency, performance, and token-revenue for developers and enterprises.
NVIDIA Unveils Rubin CPX: A New Class of GPU Designed for Massive-Context Inference
