{"data":{"id":835,"lang":"en","title":"Nvidia Launches Vera Rubin AI Platform at GTC 2026 With 336 Billion Transistors and 5x Inference Boost","summary":"Nvidia unveils Vera Rubin AI platform at GTC 2026 with 336 billion transistors and 5x inference improvement over Blackwell.","category":"technology","category_label":"Technology","published_at":"2026-03-16T15:02:58Z","updated_at":"2026-03-16T15:02:58Z","url":"https://techtools.cz/tools/news/article/835-nvidia-launches-vera-rubin-ai-platform-at-gtc-2026-with-336-billion-transistors","api_url":"https://techtools.cz/news-api/v1/articles/835?lang=en","markdown_url":"https://techtools.cz/news-api/v1/articles/835.md?lang=en","image_url":"https://blogs.nvidia.com/wp-content/uploads/2026/03/26gtc-sj-preshow-various-_MF92538_sized-scaled.jpg","word_count":186,"reading_time_minutes":1,"sources":[{"url":"https://blogs.nvidia.com/blog/gtc-2026-news/","domain":"blogs.nvidia.com"},{"url":"https://tech-insider.org/nvidia-gtc-2026-rubin-gpu-analysis/","domain":"tech-insider.org"},{"url":"https://wccftech.com/watch-nvidia-gtc-2026-ceo-jensen-huang-keynote-live/","domain":"wccftech.com"}],"translation":{"lang":"cs","id":861,"title":"Nvidia spouští platformu Vera Rubin na GTC 2026 s 336 miliardami tranzistorů a 5násobným zlepšením inference","url":"https://techtools.cz/tools/news/clanek/861-nvidia-spousti-platformu-vera-rubin-na-gtc-2026-s-336-miliardami-tranzistoru-a","api_url":"https://techtools.cz/news-api/v1/articles/861?lang=cs"},"content":"Nvidia CEO Jensen Huang kicked off GTC 2026 in San Jose on Monday before 30,000 attendees from 190 countries, officially launching the Vera Rubin platform — the successor to Blackwell. The Rubin GPU packs 336 billion transistors on TSMC's 3nm process, nearly doubling Blackwell's 208 billion, and delivers 50 petaflops of FP4 inference performance — a 5x leap over its predecessor.\n\nEach Rubin GPU carries 288GB of HBM4 memory with 22 TB/s bandwidth, nearly tripling Blackwell's 8 TB/s. The platform pairs with the new 88-core Vera CPU built on Arm v9.2 architecture and connects via NVLink 6 at 3.6 TB/s per GPU. Nvidia says the platform cuts inference token costs by 10x and reduces the number of GPUs needed to train mixture-of-experts models by 4x.\n\nRubin entered full production in early 2026, with 200,000 to 300,000 units expected this year and partner products shipping in the second half. A more powerful Rubin Ultra variant is confirmed for 2027. Nvidia also announced a multiyear partnership with Thinking Machines Lab to deploy at least one gigawatt of Vera Rubin systems for frontier model training."}}