Vera Rubin NVL72 Request capacity

Vera Rubin NVL72 / Vera Rubin availability

NVIDIA Vera Rubin NVL72 providers and availability

Where each cloud stands, how this capacity actually gets sold, and what to ask before you commit.

Updated 10 Nov 2026 · Sources linked inline

Where each provider stands

ProviderStatusWhat's publicSource
CoreWeaveCustomers liveSays Vera Rubin NVL72 is available on its cloud. Its first customer, Cognition, has run production workloads on a cluster stood up in early September 2026.coreweave.com
30 Sep 2026
Google CloudFirst instancesNVIDIA says the first A5X instance on Vera Rubin is up and running for a London startup.blogs.nvidia.com
21 Jul 2026
Microsoft AzureRacks runningNamed by NVIDIA among partners with racks running. Microsoft and Mistral plan European infrastructure with tens of thousands of GPUs.blogs.nvidia.com
21 Jul 2026
Oracle (OCI)Racks runningNamed by NVIDIA among partners with racks running, and an early adopter of co-packaged optics for the network between racks.blogs.nvidia.com
21 Jul 2026
NebiusFirst system receivedReceived its first Vera Rubin NVL72 at its Finland site and plans availability in Europe and the US.blogs.nvidia.com
21 Jul 2026
AWSAnnouncedNamed by NVIDIA at launch among the clouds expected to deploy Vera Rubin instances in 2026.nvidianews.nvidia.com
5 Jan 2026
LambdaAnnouncedNamed at launch as an NVIDIA Cloud Partner expected to deploy in 2026.nvidianews.nvidia.com
5 Jan 2026
NscaleAnnouncedNamed at launch as an NVIDIA Cloud Partner expected to deploy in 2026.nvidianews.nvidia.com
5 Jan 2026

Built from NVIDIA's and the providers' own announcements. Statuses move quickly; we recheck them and date every source.

The timeline so far

  • January 2026. NVIDIA launches the Rubin platform at CES and says partner products arrive in the second half of the year (NVIDIA).
  • May 2026. NVIDIA says Vera Rubin is ramping into full production, with shipments from the fall (NVIDIA).
  • July 2026. First racks running at CoreWeave, Google Cloud, Microsoft and Oracle; Nebius receives its first system (NVIDIA).
  • September 2026. CoreWeave announces availability, with Cognition in production (CoreWeave).

How Vera Rubin capacity is sold

Almost none of it is self-serve. The first waves go to buyers who reserved early, usually on multi-year terms with money down, against halls that are still being fitted out. If you want racks in the next few quarters you are asking for a slot in someone's delivery schedule.

That makes it a search problem. The big clouds are the visible names, but neoclouds and data center operators with allocations are often more flexible on size and start date. We ask across all of them at once. For the wider neocloud market, see Best Neocloud.

What to ask before you commit

  • Delivery date per rack, and the remedy if it slips.
  • Whole racks with the 72-GPU NVLink domain intact, or slices.
  • Fabric between racks, and whether it is non-blocking at your size.
  • Site, power source and cooling design.
  • Deposit, prepayment and what is refundable.
  • Whether you can start on GB300 or GB200 and move to Vera Rubin when it lands.

Questions

Who was first with Vera Rubin NVL72?
NVIDIA called CoreWeave the first AI cloud to bring up and validate the system (NVIDIA), and CoreWeave says Cognition is the first customer anywhere in production on it (CoreWeave).
Is Vera Rubin NVL72 generally available?
Not broadly. A few providers have racks running or first customers live. Most capacity is reserved ahead of delivery.
Can I bridge with Blackwell?
Yes, and many buyers do. Starting on GB300 NVL72 or GB200 NVL72 with a planned move to Vera Rubin is a common structure. Mention it in your request.

RackShore runs this site and may be paid a referral fee by the provider you choose. Buyers pay us nothing. Spot a mistake? Tell us.

Keep reading