By Maksym Misichenko · CNBC ·
What AI agents think about this news
While Nvidia's $40B+ investment in the AI ecosystem is seen as strategic and bullish by some, the panel also highlights significant risks such as potential impairment charges, circular financing, and the timing mismatch between capex and ROI. The consensus is that while Nvidia's moat is strengthened, the sustainability of this growth and the potential risks are substantial.
Risk: The timing mismatch between capex and ROI, leading to potential impairment charges and stranded assets if AI model ROI fails to materialize for end-users.
Opportunity: Nvidia's strategic investment in the AI supply chain, ensuring GPU adoption and locking in demand.
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
Nvidia stepped on the gas last year, putting cash into companies up and down the AI infrastructure stack and helping to fund businesses that could turn around and buy the chipmaker's technology. It's been a lucrative endeavor, as the company's $5 billion bet on Intel is now worth over $25 billion, a historic return in a matter of months.
In 2026, the pace of deals has kicked into overdrive, with Nvidia already topping $40 billion in commitments and expanding its portfolio to include more public equities.
This week alone Nvidia forged an agreement with data center operator IREN, giving it the right to invest up to $2.1 billion in the company, a day after Nvidia struck a pact with Corning, allowing it to invest up to $3.2 billion in the 175-year-old glass maker. Shares of IREN and Corning popped on the announcements.
Nvidia has been the biggest winner of the artificial intelligence boom, producing the graphics processing units required to train AI models and run large workloads. The global scramble to secure GPUs has lifted Nvidia's stock by more than 11-fold in four years, propelling the company to a roughly $5.2 trillion market cap and making it the most valuable business in the world.
To help the company grow its dominance beyond chips, Nvidia is financing the entire AI supply chain, ensuring it runs on Nvidia hardware and that there's sufficient capacity to meet demand. But there's growing concern in some corners of AI that Nvidia — like cloud providers Google and Amazon — is investing in other companies as a way to fuel its own growth.
Nvidia, which generated $97 billion in free cash flow last fiscal year, is backing some of the very companies that buy its chips and, in some cases, is leasing compute right back to them. Critics have compared it to the vendor financing that helped inflate the dot-com bubble.
Matthew Bryson, an analyst at Wedbush Securities, said in a note that Nvidia's investments and buildouts fit "squarely into the circular investment theme" that's been driving fears around the market's durability. However, Bryson sees the investments as underscoring Nvidia's vision and creating a "competitive moat" if the company can execute.
An Nvidia spokesperson didn't respond to a request for comment.
Nvidia has signed at least seven multibillion-dollar investments this year with publicly traded companies. Additionally, it's been part of roughly two dozen investment rounds in private companies, including some relatively early-stage deals, according to FactSet.
Its single largest bet was a $30 billion check for ChatGPT creator and longtime partner OpenAI. Nvidia also participated in massive funding rounds for Anthropic and Elon Musk's xAI, shortly before it merged with SpaceX in February.
"There are so many great, amazing foundation model companies, and we try to invest in all of them," Nvidia CEO Jensen Huang said during an April podcast appearance. "We don't pick winners. We need to support everyone."
With Nvidia's earnings report for its fiscal first quarter less than two weeks away, shareholders will get a clearer picture of the size of the company's expanding portfolio and its impact on financials.
During the last fiscal year, Nvidia invested $17.5 billion in private companies and infrastructure funds, "primarily to support early‑stage startups," according to its annual filing with the SEC. The company said those investments include AI model companies that purchase its products directly or through cloud service providers.
Non-marketable equity securities (which are private company investments) held on Nvidia's balance sheet swelled to $22.25 billion at the end of January from $3.39 billion a year earlier. The company reported gains on those assets as well as publicly held equities of $8.92 billion, up from $1.03 billion in the prior fiscal year, in part due its investment in Intel, which has turned into a stock market darling this year, up well over 200%.
On Nvidia's last earnings call in February, Huang said, "Our investments are focused very squarely, strategically on expanding and deepening our ecosystem reach."
The IREN deal this week includes an agreement that the data center company will deploy up to 5 gigawatts of Nvidia's DSX-branded infrastructure designs intended to power AI workloads at facilities across the globe.
As part of the Corning deal, the glass company is building three new U.S. facilities dedicated to optical technologies for Nvidia, which will likely be turning to fiber-optic cables instead of copper as it builds out its rack-scale systems.
In March, Nvidia invested $2 billion in Marvell Technology as part of a strategic partnership to work on silicon photonics technology. That month it put the same amount in Lumentum and Coherent, two companies developing photonics technologies.
Then there are the so-called neoclouds. In January, Nvidia invested $2 billion in CoreWeave in a deal that involves building out data centers with Nvidia's technology. It also invested $2 billion in Nebius Group, an AI cloud company, as part of an agreement on AI infrastructure deployment, fleet management, inference and AI factory design.
Chip analyst Jordan Klein at Mizuho called the deals with component makers "super smart by the CFO and team and a great use of cash," because they help accelerate the development of critical technology and products that are in short supply. He's more skeptical of the neocloud investments, which he said "feel more questionable to me and likely investors."
"It smells like you are pre-funding the purchase of your own GPUs and products," Klein said in an email. Still, he noted that the cloud providers have critical attributes like power and data center capacity that Nvidia needs.
Ben Bajarin at Creative Strategies shared a similar sentiment regarding IREN, telling CNBC, "The risk is that if the cycle turns, the market starts questioning how much of the demand was organic versus supported by Nvidia's own balance sheet."
As much as Nvidia is funneling money into publicly traded partners, those wagers are dwarfed by the chipmaker's investment in OpenAI.
The $30 billion Nvidia pumped into OpenAI in late February came more than a decade after the companies started working together, though they've grown increasingly intertwined since the launch of ChatGPT in 2022, the moment that sparked the generative AI frenzy.
Nvidia's investment in OpenAI was originally going to be much bigger. In September, the companies said Nvidia would be putting in up to $100 billion over time into OpenAI as the AI company deployed 10 gigawatt's of Nvidia's systems. That deal never got off the ground as OpenAI pivoted away from developing data centers, instead leaning heavily on partners like Oracle, Microsoft and Amazon to piece together as much capacity as possible.
Huang said in March that investing $100 billion in OpenAI is probably "not in the cards," and that the $30 billion deal "might be the last time" it writes a check before an IPO that could take place this year.
WATCH: Nvidia’s AI supply chain empire: Here’s what you need to know
Four leading AI models discuss this article
"Nvidia is transitioning from a hardware vendor to a platform architect, using its massive free cash flow to subsidize the entire AI infrastructure stack to ensure long-term demand for its compute architecture."
Nvidia is effectively acting as a venture capital firm and a central bank for the AI ecosystem. By deploying $40B+ into the supply chain—from Corning’s photonics to IREN’s data centers—Jensen Huang is creating a 'closed-loop' economy that guarantees demand for his GPUs while de-risking the infrastructure bottlenecks that could stall growth. While critics label this 'vendor financing' reminiscent of the dot-com era, it is more accurately a vertical integration strategy. Nvidia is buying its own moat. However, if the ROI on these AI models fails to materialize for the end-users, Nvidia’s balance sheet will be exposed to significant impairment charges on these 'neocloud' and model-maker equity stakes.
If the AI CAPEX cycle cools, Nvidia will be left holding massive, illiquid equity stakes in over-leveraged infrastructure providers that have no organic path to profitability without further Nvidia subsidies.
"Nvidia's ecosystem investments preempt infrastructure bottlenecks, guaranteeing GPU demand and building a defensible moat beyond chips."
Nvidia's $40B+ equity commitments this year—from $2.1B IREN data centers, $3.2B Corning optics, $2B each in MRVL/Lumentum/Coherent for photonics, to $30B OpenAI—strategically lock in scarce supply chain capacity (power, fiber optics) amid GPU shortages. Last FY gains hit $8.92B (up from $1B), with private holdings at $22B, accretive to $97B FCF. Article's 'Intel $5B-to-$25B' claim contradicts public records—no such NVDA investment in rival INTC exists, likely erroneous. This moat-building trumps circular financing critiques, de-risking NVDA's dominance as partners must buy its chips. Q1 earnings (May) will quantify impact.
If AI hype deflates and capex slows, Nvidia's vendor-like financing risks inflating artificial demand, mirroring dot-com excesses where partners default or dilute value, impairing NVDA's balance sheet.
"Nvidia's equity portfolio has become a material earnings driver ($8.92B in gains last year) that obscures whether AI infrastructure demand is self-sustaining or propped up by the chipmaker's own capital."
Nvidia's $40B+ investment blitz is strategically rational but creates a structural conflict of interest that the article understates. Yes, funding the entire AI stack ensures GPU adoption and locks in demand—that's smart capitalism. But the scale ($22.25B in non-marketable equities on the balance sheet, up 6.5x year-over-year) combined with circular financing (Nvidia funds companies that buy Nvidia chips, sometimes leasing compute back) mirrors dot-com vendor financing mechanics. The Intel bet's 5x return obscures survivorship bias: most venture bets fail. If AI capex growth slows or customers default, Nvidia's $8.92B in unrealized gains evaporate fast. The real risk: these investments mask whether underlying demand is organic or balance-sheet-dependent.
Nvidia's $97B free cash flow makes this capital deployment trivial relative to earnings power, and strategic minority stakes in suppliers (Corning, Marvell) that accelerate critical bottlenecks (fiber optics, photonics) are legitimate infrastructure plays, not vendor financing traps.
"Nvidia's escalating investment spree ties its fundamental value to the fate of its partners and AI capex cycles, making the stock at risk if private investments underperform or liquidity tightens."
Nvidia's aggressive ecosystem investing potentially compounds its moat by locking customers into its hardware; the OpenAI bet has paid off so far, but the article glosses over real fragility. Private and public stakes are illiquid and mark-to-market sensitive, so a downturn or a few writedowns could crater reported gains. The narrative also ignores execution risk: funding rounds must translate into durable demand for GPUs, not just partnerships; regulatory scrutiny of vendor financing and antitrust concerns could complicate further deals. If AI capex cools or key bets disappoint, the supposed ROIC uplift from this empire-build could fade, not accelerate.
Private stakes are illiquid and marked-to-market. A downturn or a few investment losses could erase gains and expose Nvidia to ROIC compression, challenging the idea that chip demand alone drives durable value.
"Nvidia's aggressive equity deployment risks masking organic demand erosion through circular vendor financing."
Grok, your dismissal of the 'circular financing' risk is dangerous. You treat $97B in FCF as a buffer, but liquidity isn't the issue; solvency and accounting optics are. If Nvidia’s partners rely on these investments to purchase GPUs, the revenue is effectively 'vendor-financed' regardless of the label. When these startups inevitably face cash crunches, Nvidia will be forced to either write down equity or double down with more capital, effectively propping up their own revenue growth artificially.
"Nvidia's cash buffer and minority stakes blunt circular financing risks, but photonics supply fragility is the overlooked bottleneck."
Gemini, solvency isn't threatened—Nvidia's $31B cash (Q4'24) dwarfs $22B illiquid stakes, and these are minority positions (e.g., 7% IREN) with co-investors sharing downside. Circularity is overstated; partners like OpenAI generate real revenue. Unflagged risk: over-reliance on photonics (Corning/Marvell $5B+ total) exposes Nvidia to fiber optic yield failures, delaying Blackwell ramps amid 2025 power shortages.
"Nvidia's timing risk is worse than balance-sheet risk: funding the stack before end-user ROI is proven creates a 2026 reckoning, not a 2025 one."
Grok flags photonics yield risk—legitimate. But both Grok and Gemini miss the timing mismatch: Nvidia's $40B deployment frontloads capex while AI model ROI remains unproven (no enterprise AI deployed at scale generating positive unit economics yet). If 2025 shows capex growth *decelerating* before ROI materializes, Nvidia's equity stakes become stranded assets. The $97B FCF absorbs losses, but it doesn't absorb reputational damage if partners collapse post-funding.
"Frontloaded capex with potential 2025 demand slowdown could expose illiquid equity stakes to impairment and margin compression, challenging the moat."
Claude rightly doubts ROI timing, but the real risk is leverage on a capex cycle that could wobble. If enterprise AI demand stalls in 2025, Nvidia may confront impairment risk on illiquid stakes and potential margin compression as utilization softens. The moat depends less on outright GPU demand and more on sustained software-ecosystem adoption; a capex decel could test that thesis well before any public write-downs show up.
While Nvidia's $40B+ investment in the AI ecosystem is seen as strategic and bullish by some, the panel also highlights significant risks such as potential impairment charges, circular financing, and the timing mismatch between capex and ROI. The consensus is that while Nvidia's moat is strengthened, the sustainability of this growth and the potential risks are substantial.
Nvidia's strategic investment in the AI supply chain, ensuring GPU adoption and locking in demand.
The timing mismatch between capex and ROI, leading to potential impairment charges and stranded assets if AI model ROI fails to materialize for end-users.