NVIDIA at Google Cloud Next ’26 : US Pioneer Global VC DIFCHQ SFO NYC Singapore – Riyadh Swiss Our Mind

Google Cloud Next ’26, Google Cloud and NVIDIA announced a deep expansion of their partnership, focusing on co-engineered AI infrastructure designed for agentic AI, mixture-of-experts (MoE) models, and physical AI

Key announcements include the adoption of next-generation NVIDIA Blackwell and Rubin platforms, expanded GPU-powered virtual machine options, and new software integrations to improve AI performance and reduce costs.

NVIDIANVIDIA +3
Key Infrastructure and Hardware Announcements
  • NVIDIA Blackwell (B200/GB200) Adoption:Google Cloud is expanding its AI Hypercomputer with A4 and A4X VMspowered by NVIDIA Blackwell GPUs, offeringmassive scale for training and inference.
  • NVIDIA Rubin Platform: Google announced it will be among the first cloud providers to deploy NVIDIA Vera Rubin-based instances (specifically the Rubin NVL72) in the second half of 2026.
  • G4 VMs with RTX PRO 6000 Blackwell:Generally available to accelerate AI inference, 3D simulation, and agentic AI workloads.
  • DGX Cloud on Google Cloud: NVIDIA DGX Cloud with GB200 NVL72 will be available on Google Cloud in 2025 to provide a serverless AI supercomputing experience.
    Google CloudGoogle Cloud +3
AI Software and Service Integrations
  • NVIDIA NIM & NeMo: NVIDIA NIM inference microservices and NeMo framework for developing custom models are now integrated into Google Cloud services, including GKE and Cloud Run.
  • NVIDIA Omniverse on GCP: NVIDIA Omniverse and Isaac Sim are available on the Google Cloud Marketplace to support industrial digitalization, digital twins, and robotics development.
  • GKE Inference Gateway: A new tool, integrated with NVIDIA Triton Inference Server and NeMo Guardrails, designed to reduce AI serving costs by up to 30% and improve throughput.
  • Data Science Boost: Google Cloud’s Vertex AI now supports NVIDIA H100 GPUs and integrated Triton Inference Server for faster development and deployment.
    Google CloudGoogle Cloud +4
Joint Initiatives
  • Public Sector AI Accelerator: Google Public Sector and NVIDIA are launching a year-long accelerator program for Independent Software Vendors (ISVs) developing AI solutions for public sector applications.
  • Confidential Computing: Expansion of confidential virtual machines (VMs) to support NVIDIA H100 GPUs, allowing secure processing of sensitive data.
  • Startup Support: Qualified NVIDIA Inception members can receive up to $350,000 in Google Cloud credits to accelerate AI development.
    Google CloudGoogle Cloud +2
These collaborations aim to enhance the performance of generative AI and agentic AI workloads, with a focus on lowering total cost of ownership (TCO) for customers running large-scale AI projects
https://www.nvidia.com/en-us/events/google-cloud-next/