Explained: What Are AI Neoclouds And How Nvidia's Revenue-Sharing Model Has Fueled Its Rise

Nvidia is guaranteeing minimum revenue on the GPU capacity it sells to neoclouds, a backstop that is quietly reshaping how the AI infrastructure boom gets funded, according to Goldman Sachs.

Advertisement
Read Time: 4 mins
Quick Read
Summary is AI-generated, newsroom-reviewed
  • Neoclouds rent GPU compute for AI training, focusing on few large customers, not broad enterprises
  • Nvidia offers revenue guarantees to neoclouds, sharing risk and profit from GPU capacity
  • Indian neoclouds use Nvidia’s revenue-share model to ease financing amid a sovereign push in AI infrastructure
Did our AI summary help?
Let us know.

A new category of cloud company has emerged at the centre of the AI infrastructure buildout: the neocloud. Unlike legacy hyperscalers such as Amazon Web Services, Microsoft Azure or Google Cloud, neoclouds are built from the ground up to do one thing — rent out GPU compute for AI training and inference, at scale, and often to a small number of large customers rather than a broad enterprise base.

Firms such as CoreWeave, Nebius, Firmus and Sharon AI fall into this bucket. Many are backed directly or indirectly by chipmakers, and several operate under sovereign or state-linked structures rather than as pure private enterprises.

Advertisement

What has changed in recent months, according to a Goldman Sachs research note dated 31 July 2026, is how these neoclouds get their GPUs funded in the first place.

The Backstop Model, Explained

Nvidia has started offering select neocloud customers a minimum revenue guarantee on the GPU capacity it supplies. If a neocloud is unable to lease out that capacity to end customers, Nvidia effectively backstops the shortfall. In exchange, Nvidia takes a share of the neocloud's revenue above a certain threshold.

Advertisement

ALSO READ: One Founder, AI Workforce: The New Formula For Million-Dollar Startups

The mechanics matter because they lower the risk of building out GPU capacity ahead of confirmed demand. A neocloud no longer has to bet entirely on its own ability to fill data centre capacity before the model works financially. Nvidia is effectively underwriting a portion of that risk, in return for upside participation once utilisation climbs.

Goldman Sachs analysts Katherine Murphy and Michael Ng named sovereign neoclouds Firmus and Sharon AI as two of the earliest partners in this revenue-sharing and credit-support arrangement.

Advertisement

Firmus: 170,000 GPUs, Indonesia

Firmus has partnered with Nvidia to build a 360-megawatt AI factory campus in Batam, Indonesia, housing 170,000 Nvidia accelerators spanning the Grace-Blackwell, Vera-Rubin and Vera chip generations. Under the arrangement, Firmus sells Nvidia-powered cloud services to end customers, while Nvidia earns both standard product revenue on the chips and a share of the cloud revenue generated from the supported capacity. The campus is expected to come online between 2027 and 2028.

Sharon AI: A $1.32 Billion Contract

Sharon AI, an Australian neocloud, has signed a five-year, $1.32 billion cloud computing contract with an undisclosed AI lab. It is deploying its first AI Factory across New Zealand data centre infrastructure, with 132 megawatts of total capacity — 116 megawatts already contracted — and more than 62,000 Nvidia GPUs expected to be deployed by mid-2027. The build includes up to 40,000 Grace Blackwell GPUs and a 600 petabyte storage agreement with Vast Data. Revenue is expected to begin flowing in the first half of 2027.

Why It Matters Beyond Nvidia

Goldman Sachs views the continued strength in neocloud demand as a positive read-through for server vendors that supply the physical infrastructure behind this build-out, naming Dell and Super Micro Computer (SMCI) as key beneficiaries. The logic is straightforward: every GPU that Nvidia backstops still needs to sit inside a server rack, and neoclouds buying at this scale are large, repeat customers for server makers.

The broader signal is that AI infrastructure spending is increasingly being de-risked through vendor financing arrangements rather than pure market demand, a dynamic worth watching as more chipmakers and cloud investors look to replicate the model.

Advertisement

India & Neoclouds

The same playbook is now showing up closer to home. Indian neoclouds such as NeevCloud and Utho Cloud say Nvidia's revenue-share backstop is already easing the financing burden of their own GPU build-outs, alongside a broader sovereign push from Yotta, L&T and E2E Networks.

The Indian players involved in neocloud building
Photo Credit: NDTV Profit

ALSO READ: Indian IT's AI Story Is Changing, But BofA Flags A New Earnings Risk

Essential Business Intelligence, Sharp Market Insights, Practical Personal Finance Advice, Daily Fuel, Gold and Silver Prices and Latest Stories — On NDTV Profit.


Loading...