NVIDIA Blackwell GPUs are coming soon to Lambda!

• NVIDIA GB200 Grace™ Blackwell Superchip

  GB200 NVL72 provides up to 30X faster real-time LLM inference and 4X faster training performance compared to the NVIDIA Hopper architecture.
• NVIDIA HGX B200

  Up to 15X more real-time inference and 3X faster training compared to the NVIDIA Hopper architecture generation.