Seonglae ChoRubin GPUs are built on TSMC’s 3 nm process, use HBM4 memory and are slated for mass production in late 2025 with availability in early 2026; they deliver roughly 50 petaflops of FP4 compute (versus 20 petaflops in Blackwell) and about 1.2 exaflops of FP8 training performance, thanks to enhanced NVLink and memory bandwidth
NVIDIA Kicks Off the Next Generation of AI With Rubin — Six New Chips, One Incredible AI Supercomputer
NVIDIA today kickstarted the next generation of AI with the launch of the NVIDIA Rubin platform, comprising six new chips designed to deliver one incredible AI supercomputer.
https://nvidianews.nvidia.com/news/rubin-platform-ai-supercomputer
Nvidia announces new AI chips months after latest launch as market competition heats up
Nvidia CEO Jensen Huang announced the new AI chip architecture, dubbed "Rubin," ahead of the COMPUTEX tech conference in Taipei.
https://www.cnbc.com/2024/06/02/nvidia-next-generation-ai-chips-rubin-blackwell.html
