Google Cloud and NVIDIA Bring Vera Rubin A5X Instances to General Availability

What A5X combines

According to NVIDIA, A5X instances bring together NVIDIA Vera Rubin NVL72, NVIDIA ConnectX-9 SuperNICs, Google Cloud Virgo Network and Google Cloud's fourth-generation advanced liquid-cooling technology. The target is demanding agentic inference and reinforcement learning workloads at scale. The platform is meant to give cloud customers an integrated system for training, inference and large-scale reinforcement learning.

NVIDIA says A5X brings Vera Rubin into Google Cloud's AI Hypercomputer as an integrated system, combining compute, networking and storage. The idea is to support learning through interaction, which requires a continuous cycle of generating experience, evaluating outcomes and updating models.

First customers: Ineffable Intelligence and WPP

Ineffable Intelligence, a London-based technology company founded by AlphaGo creator David Silver, was the first customer to access A5X. The company is using the platform to advance large-scale reinforcement learning and to shape "superlearner" systems that learn continuously from experience.

Silver, founder and CEO of Ineffable Intelligence, said in NVIDIA's announcement: "We're building superlearners that continually improve through their own experience." He added that this demands substantial compute and a tight connection between generating experience and training on it, and pointed to the high-bandwidth connectivity between the Vera CPU and Rubin GPUs and between the Rubin GPUs within the rack.

WPP, a London-based marketing and communications company, has demonstrated its creative agent reference architecture on Vera Rubin-based A5X instances. It used NVIDIA Nemotron models and NVIDIA OpenShell, part of the NVIDIA OpenShell Reference Architecture for safe AI agents. The agents turn conversational requests into multistep creative workflows, such as video generation, upscaling and restyling, with human approval before execution. OpenShell provides governed sandboxes for agent execution, helping control access to sensitive brand assets and creative tools. In WPP's initial tests, workloads that typically required a multi-node cluster ran on a single A5X instance, according to NVIDIA.

Confidential computing and power efficiency

The announcement also covers data protection. NVIDIA notes that organizations deploying agentic systems need to protect proprietary models and sensitive data, even on infrastructure they do not own. Google Cloud will be among the first cloud providers to offer NVIDIA Confidential Computing at rack scale on bare metal instances such as A4X Max instances powered by NVIDIA GB300 NVL72.

Google Cloud is also making Confidential Computing generally available on G4 VMs powered by NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs. This is positioned as a flexible platform for confidential inference and fine-tuning. Vertebrae, Polaris Cloud and Confidential Core AI are using Confidential G4 for conversational privacy, verifiable agent execution and policy-governed enterprise intelligence.

On energy, Google Cloud and NVIDIA are exploring NVIDIA DSX MaxLPS capabilities for Vera Rubin-based infrastructure. The aim is to help optimize performance per watt and expand capacity within available data-center power budgets.

What it means for customers

For cloud customers, the news means a Vera Rubin-based option for agentic inference and reinforcement learning is due on Google Cloud next month, with European availability planned. The stated scale of up to 960,000 Rubin GPUs across multiple sites signals the capacity Google Cloud intends to build. Organizations with sensitive workloads also gain confidential computing options on G4 VMs, with rack-scale support planned on bare metal instances such as A4X Max.