Nvidia says Google Cloud’s A5X instances using its Vera Rubin platform will become generally available next month. Google Cloud is also making Confidential Computing generally available on G4 virtual machines for confidential inference and fine-tuning.
A5X is part of Google Cloud’s AI Hypercomputer. According to Nvidia, it combines compute, networking and storage to support training, inference and large-scale reinforcement learning. Google Cloud plans to scale its Vera Rubin infrastructure to as many as 960,000 Rubin GPUs across multiple sites.
Ineffable Intelligence, a London-based company founded by AlphaGo creator David Silver, was the first customer to access A5X. Nvidia says it is using the platform for large-scale reinforcement learning research and work on systems that learn continuously from experience.
WPP has also demonstrated a creative agent setup on A5X using Nvidia Nemotron models and OpenShell. According to Nvidia, WPP’s initial tests ran workloads that typically required a multi-node cluster on one A5X instance.
Read nextGoogle Cloud adds OpenAI models to its model gardenGoogle Cloud expands confidential computing
The G4 virtual machines use Nvidia RTX PRO 6000 Blackwell Server Edition GPUs. Nvidia says Confidential Computing on G4 provides a platform for confidential inference and fine-tuning.
Google Cloud also plans to offer Nvidia Confidential Computing at rack scale on bare metal instances such as A4X Max, which use Nvidia GB300 NVL72. The companies are exploring Nvidia DSX MaxLPS capabilities for Vera Rubin infrastructure to improve performance per watt and fit more capacity within available data-center power budgets.
A5X instances are scheduled for general availability next month.
