Generative AI Computing Platform

NVIDIA AI - One Architecture | Train and Deploy Everywhere From Two-Year Rhythm to One-Year Rhythm | Training & Inference | x86 & Arm | Hyperscale & Enterprise 2021 2023 2024 2025 GH200NVL GB200NVL GX200NVL Arm Training & Inference GH200 GB200 GX200 Arm Inference A100 H100 H200 B100 X100 X86 Training & Inference L4OS B40 X40 X86 Enterprise & Inference Quantum 400G 800G 1,600G InfiniBand Al Infrastructure Spectrum-X 400G 800G 1,600G Ethernet Enterprise & Hyperscale Al Infrastructure 26 NVIDIA.

View entire presentation