Nvidia Blackwell B200 GPU

30 times performance increase for LLM inference workloads and can scale to AI models with up to 10 trillion parameters

The Nvidia Blackwell B200 GPU, hailed as the world's most powerful chip for AI, delivers a groundbreaking 30x performance boost for large language model (LLM) inference, supporting models with up to 10 trillion parameters. Designed for the new era of computing, its GB200 NVL72 configuration combines 36 Grace CPUs and 72 Blackwell GPUs in a liquid-cooled, rack-scale design, functioning as a single massive GPU. This setup enables real-time trillion-parameter LLM inference at unprecedented speeds. With 40x faster AI inference than previous platforms, 40 TB of high-speed memory, and advanced networking integration, the Blackwell B200 excels in complex workloads, from reasoning systems to real-time video generation. Its energy efficiency and scalability make it a game-changer for AI-driven applications.