Nvidia's Growth Story Now Hinges on Three Unproven Bets

Published on 10/28/2026 at 11:11 | Editorial boerse-global.de

Nvidia floated a $2,000 per-GPU annual software fee and logged over 3 million Chinese GPU orders it may not fill, alongside a 2.5-fold substrate capacity plan.

Nvidias, Growth, Story, Now, Hinges, Three, Unproven, Bets, Nvidia, Chinese Illustration mit AI erstellt.

Nvidia has spent the past week sketching out the demand side of its AI ambitions, but the more consequential disclosures concern whether the company can actually deliver. Between a proposed per-GPU software fee, a wave of Chinese orders it cannot fully fill, and a multi-partner manufacturing build-out, the chipmaker is asking investors to weigh intentions against booked business.

At the GTC Berlin event on October 22, CEO Jensen Huang floated licensing the company's agent software suite at roughly $2,000 per GPU per year, according to media reports. The figure represents a potential recurring revenue stream tied directly to hardware deployment — not a signed contract or reported software income. How widely customers would adopt such an offering is what would determine its economic weight.

That pricing idea sits alongside the OpenShell Reference Architecture unveiled on October 21. Nvidia describes the framework as a way to deploy and manage AI agents across enterprise infrastructure and edge systems, and says it is available immediately, with partners spanning cloud, software, cybersecurity and systems. The distinction matters: the architecture is a shipping product, while Huang's licensing remark is a possible monetization approach. Availability alone says nothing about commercial traction.

Orders Outpace the Factory Floor

Demand signals are arriving faster than supply. Reuters reported on October 21 that ByteDance, Alibaba and Tencent placed orders for more than 3 million RTX PRO 5500 graphics cards, with a combined list value above $35 billion. Nvidia could not guarantee it would fill the orders in full, according to the report, because demand outstripped planned production. Those bookings therefore do not translate into realized revenue — the gap between order interest and shipped hardware is where the valuation debate now sits.

Should investors sell immediately? Or is it worth buying Nvidia?

The company moved on the supply side the same day, announcing a capacity expansion with several suppliers. Substrate agreements are designed to lift capacity 2.5-fold over three years, with Amkor, IBIDEN, KYEC, Samsung Electro-Mechanics, SPIL and Unimicron among the contributors. The measures cover Nvidia's accelerated computing platforms and full-stack AI infrastructure, addressing several links in the supply chain at once rather than relying on a single partner. These are planned expansions, not additional capacity already online.

Nvidia is also putting money behind the plan: a $1.1 billion prepayment to KYEC to reserve capacity, including new test facilities in the United States. That outlay is a capacity reservation, not revenue from AI system sales — spending and the growth it is meant to enable sit on opposite sides of the ledger.

Customers Line Up for the Rollout

The demand context arrived a day earlier. Nvidia announced expanded collaboration with Meta's Muse models on October 21, with Meta intending to broaden its use of Vera Rubin and Vera CPUs as well as Spectrum-X and BlueField-4. Nvidia simultaneously disclosed a $3 billion investment in Meta's equity offering. In parallel, Nvidia and Oracle laid out plans to scale OCI superclusters to multi-gigawatt size, with Vera Rubin NVL72 and Vera CPUs now generally available on OCI.

Nvidia at a turning point? This analysis reveals what investors need to know now.

Those deals explain why the manufacturing push matters strategically, though they substitute for neither confirmed results nor delivered capacity. The thread running through all of it: Nvidia is pairing wider adoption of its platforms with long-term capacity security, while a software licensing model waits in the wings. Whether that turns into commensurate growth depends on execution — on filling orders, bringing announced capacity online, and converting a proposed fee into paying users. License price, order book and expansion plans mark separate stages of the same journey, and none of them is finished.

Nvidia Stock: New Analysis - 28 October

Fresh Nvidia information released. What's the impact for investors? Our latest independent report examines recent figures and market trends.

Read our updated Nvidia analysis...