Deyron Labs graphic: Google Cloud A5X on Vera Rubin NVL72; general availability next month, up to 960,000 Rubin GPUs planned, expansion to Europe planned

What happened

In a post dated 21 October 2026, NVIDIA announced with Google Cloud that A5X instances, “accelerated by NVIDIA Vera Rubin NVL72 systems”, are heading to general availability. The post says availability of the instances “is coming next month, with planned expansion of availability to Europe”. We read NVIDIA’s post; we did not read a Google Cloud announcement, which NVIDIA links.

Key details

The details below come from NVIDIA’s post.

  • Hardware: A5X combines Vera Rubin NVL72, NVIDIA ConnectX-9 SuperNICs, Google Cloud’s Virgo Network and its fourth-generation liquid cooling.
  • Scale: Google Cloud plans to scale to “as many as 960,000 Rubin GPUs across multiple sites”. This is a plan, not a delivered figure.
  • Customers: Ineffable Intelligence, a London company founded by David Silver, was the first customer to access A5X. WPP demonstrated a creative-agent reference architecture on A5X using NVIDIA Nemotron models; NVIDIA says that in WPP’s initial tests, workloads that typically needed a multi-node cluster ran on a single A5X instance.
  • Confidential computing: rack-scale confidential computing is offered on bare-metal instances such as A4X Max (GB300 NVL72), and confidential G4 VMs with RTX PRO 6000 Blackwell Server Edition are generally available.
  • Not stated: the post gives no prices, regions beyond the planned European expansion, or exact launch date.

Why it matters

For teams that rent rather than build AI capacity, this puts NVIDIA’s next rack-scale generation on a major cloud’s roadmap, with a stated timeline of about a month. The announcement is from NVIDIA, so scale, performance and the WPP result are the companies’ claims and no independent measurements are given; pricing and regional availability are what customers will need to check at launch.

Disclosure: Claude, made by Anthropic, was one of the AI tools used to research and draft this article. Details come from NVIDIA’s post of 21 October 2026, read on 22 October 2026.

Sources

  1. Google Cloud and NVIDIA Announce General Availability of NVIDIA Vera Rubin Infrastructure Primary source , NVIDIA