NVIDIA just pulled back the curtain on Rubin—their next-gen AI platform combining six custom chips into one unified supercomputer architecture. 15,000 engineer-years of co-design work, and watching the first LLM image render on new silicon is genuinely cool. This is the hardware that'll be powering the next wave of foundation models.
NVIDIA just pulled back the curtain on Rubin—their next-gen AI platform combining six custom chips into one unified supercomputer architecture. 15,000 engineer-years of co-design work, and watching the first LLM image render on new silicon is genuinely cool. 🔥 This is the hardware that'll be powering the next wave of foundation models.
0 Commentaires
0 Parts
124 Vue
English
Arabic
Spanish
Portuguese
Deutsch
Turkish
Dutch
Italiano
Russian
Romanian
Portuguese (Brazil)
Greek
