Nvidia Rewrites the Rules: From Chip Seller to AI Infrastructure Financier as Rubin Ultra Stays on Track
Published on 07/13/2026 at 13:42 | Redaktion boerse-global.de
Nvidia is quietly remaking itself from a hardware vendor into a financial engine for the artificial-intelligence economy. Since the start of July, the company has been leaning hard into a strategy it calls “AI Factories,” where it uses its own balance sheet to extend credit lines and take revenue-sharing stakes in smaller cloud providers. The shift marks a profound departure from the pure-chip-supplier model that built Nvidia’s dominance, and it is already reshaping how investors think about the stock.
The move is driven by necessity. The biggest hyperscalers — Microsoft, Google, Amazon — have largely finished their initial buildouts, so the next wave of AI demand is coming from startups, national labs, and government cloud projects that lack deep pockets. Nvidia, with a market capitalisation of roughly $4.3 trillion, is effectively banking its own customers. One example: Sharon AI is now deploying up to 40,000 Grace-Blackwell GB300 processors, while partner Firmus is building a 360-megawatt AI campus in Indonesia. The goal is to lock in recurring, usage-based revenue rather than relying on one-off hardware sales.
Management has been at pains to stress that the product roadmap remains solid alongside this business-model overhaul. At a Morgan Stanley investor conference in California, CEO Jensen Huang and CFO Colette Kress publicly pushed back against rumours that the next-generation “Rubin Ultra” platform would slip from its 2027 target to 2028. They acknowledged some component tweaks — the original “Kyber” rack is being replaced with an optimised version for larger supercomputing clusters — but insisted the core delivery dates are unchanged. Morgan Stanley analyst Joseph Moore reiterated his “Overweight” rating after the meeting, highlighting Nvidia’s privileged access to scarce DRAM capacity as a competitive moat.
The architecture that bridges the present and future is the Vera-Rubin line, which will follow the current Blackwell generation. Volume production of Rubin is slated to begin in the second half of 2026, with an emphasis on efficiency per watt rather than raw peak performance. That shift aligns with what Nvidia calls the “year of inference” — the realisation that 80 to 90 per cent of data-centre compute loads are now dedicated to generating outputs from trained models, not training the models themselves. New cooling systems using closed-loop water circuits are designed to cut local water consumption to near zero, a nod to the growing environmental constraints on large-scale AI facilities. Beyond Rubin Ultra, the “Feynman” platform is already pencilled in for 2028.
Should investors sell immediately? Or is it worth buying Nvidia?
The stock itself has been reflecting a market that is cautiously optimistic but not yet exuberant. After a strong week that saw the shares gain roughly 8 per cent, Nvidia closed on Friday at €184.60 on the Xetra exchange. Monday brought a modest pullback of 1.5 per cent to €181.78, leaving the price about 10 per cent below its 52-week high of €202.50 set on 14 May. Technically, the stock is clinging to its 50-day moving average of €181.46, a level many traders view as a near-term support. It sits comfortably above the 200-day average of €164.78, and the relative strength index of 58.6 suggests there is headroom before the stock becomes overbought. Year to date, Nvidia is up about 13 per cent, and it has rallied nearly 30 per cent from the July 2025 low of €140.62.
Analysts see more runway ahead. The average price target stands at €264.16, implying potential upside of 43 per cent from current levels. That optimism is not pinned purely on chip sales; it also reflects bets on “sovereign AI” projects — national cloud initiatives in India, Brazil, and the United States that could become the next major growth engine.
Yet the new financing model introduces risks that Nvidia has not carried before. By extending credit to smaller cloud operators, the company now bears not just the risk of an order slowdown but the credit risk of its partners. If those customers struggle, Nvidia’s own balance sheet could come under pressure. On the supply side, an external risk surfaced on 9 July when Taiwan’s central bank governor, Yang Chin-long, warned of speculative overbuilding and excessive borrowing in the AI sector. With TSMC serving as Nvidia’s primary foundry, institutional investors are watching credit and debt trends in Taiwan’s technology ecosystem with growing attention.
Nvidia at a turning point? This analysis reveals what investors need to know now.
For now, the combination of a confirmed product cadence, a broadening customer base (roughly 20 per cent of demand now comes from labs, governments, and enterprises outside the hyperscalers), and a stable gross margin in the mid-70 per cent range gives Nvidia a strong hand. But the warning from Taipei is a reminder that, as Nvidia becomes more deeply embedded in the financial infrastructure of AI, its risks are no longer confined to Silicon Valley.
Ad
Nvidia Stock: New Analysis - 13 July
Fresh Nvidia information released. What's the impact for investors? Our latest independent report examines recent figures and market trends.
Disclaimer regarding our articles: No investment advice, no buy or sell recommendation. Information on prices, companies, and markets is provided without guarantee; changes are possible at any time. Stock market transactions can lead to substantial losses. Our articles are created and reviewed in whole or in part automatically with the support of AI.
