Nvidia’s transformation from chip supplier to AI ecosystem architect reached a new milestone this month. The company’s equity investments in AI-related companies have topped $40 billion, a figure that places it alongside sovereign wealth funds and the largest institutional investors in terms of sheer capital deployed into the artificial intelligence sector.

But the headline number only tells part of the story. It is the composition of the portfolio — and the returns already materializing — that reveals how deliberately Nvidia has constructed its investment strategy. The crown jewel is a $5 billion stake in Intel acquired in September 2025 that has since quintupled in value to more than $25 billion, turning what many viewed as a surprising bet into one of the most profitable equity investments in recent semiconductor history.

The Numbers

Nvidia’s investment portfolio has grown rapidly throughout 2025 and into 2026. As we detailed in our earlier breakdown of the $40 billion commitment, the portfolio spans multiple layers of the AI stack — from foundational model companies like OpenAI (~$30 billion committed) to infrastructure players like Corning ($3.2 billion) and IREN ($2.1 billion), plus $17.5 billion deployed across private AI startups and infrastructure funds according to SEC filings.

What has changed since then is the performance of the portfolio. The Intel stake, in particular, has emerged as a case study in how strategic investing can generate both financial returns and commercial leverage simultaneously.

Here is a snapshot of the key positions as of May 2026:

  • OpenAI: ~$30 billion committed, making Nvidia one of the largest investors in the ChatGPT maker alongside Microsoft and SoftBank
  • Intel: ~215 million shares acquired at $23.28 per share ($5 billion initial investment), now valued at over $25 billion after Intel’s stock surge
  • Corning: Up to $3.2 billion in the fiber optics and specialty materials company
  • IREN: Up to $2.1 billion in the AI-focused data center operator
  • Private companies and infrastructure funds: $17.5 billion across early-stage AI ecosystem companies

The aggregate portfolio value, accounting for unrealized gains on positions like Intel and estimated appreciation in the private holdings, likely exceeds the $40 billion committed capital figure by a substantial margin.

The Intel Partnership

When Nvidia acquired roughly 215 million shares of Intel at $23.28 per share in September 2025, the deal raised eyebrows. Nvidia and Intel had been competitors for decades — in GPUs, in data center processors, in the broader battle over how compute architectures would evolve in the AI era. Taking a major equity stake in a rival was, at minimum, unconventional.

The FTC approved the transaction after reviewing it for potential antitrust concerns, clearing the way for what has become one of the most consequential corporate partnerships in the semiconductor industry.

The strategic logic has become clearer over the ensuing months. Rather than viewing Intel purely as a competitor to be defeated, Nvidia under CEO Jensen Huang appears to have identified Intel’s manufacturing capabilities and x86 architecture expertise as complementary assets — tools that could extend Nvidia’s reach into market segments where its own GPU-centric architecture is not the natural fit.

On May 11, Intel CEO Lip-Bu Tan provided the most concrete evidence yet of what this partnership is producing. In a public appearance, Tan teased what he called “exciting new products” developed jointly with Nvidia, hinting at a collaboration that goes far deeper than a passive equity investment.

The Product Roadmap

Based on disclosures from both companies, the Nvidia-Intel collaboration is producing at least two distinct product categories:

NVIDIA-custom x86 CPUs for data centers. This is perhaps the most significant development. Nvidia is working with Intel to design custom x86 processors tailored specifically for data center environments where Nvidia GPUs handle the AI workloads. The logic is straightforward: AI inference and training clusters need CPUs to handle orchestration, networking, and non-GPU compute tasks. Today, many of those CPUs come from AMD or Intel without Nvidia’s direct involvement. By co-designing x86 CPUs with Intel, Nvidia can optimize the entire server stack — GPU plus CPU — for its own workloads, potentially improving performance and reducing costs for its data center customers.

x86 SOCs with integrated NVIDIA RTX GPU chiplets for PCs. The second product line targets the client computing market. These system-on-chip designs would combine Intel’s x86 CPU cores with Nvidia’s RTX GPU chiplets in a single package, creating integrated processors for laptops and desktops that blend the strengths of both companies. If these products reach market, they would represent a direct challenge to AMD’s Ryzen AI processors, which integrate AMD CPU and GPU cores in a similar fashion, and to Apple’s M-series chips, which combine CPU and GPU on a single die.

The chiplet approach is notable because it allows each company to contribute its best silicon without requiring either to redesign its core architecture. Intel fabricates the x86 CPU chiplets using its process technology, Nvidia contributes RTX GPU chiplets manufactured at TSMC or another foundry, and the two are connected via advanced packaging — a division of labor that plays to each company’s strengths.

What the Market Saw

Intel shares climbed following the May 11 product announcements, extending a rally that has been building since the Apple-Intel chip manufacturing deal disclosed the same week. For investors who bought Intel at its 2025 lows, the stock’s recovery has been dramatic. For Nvidia, which acquired its shares at $23.28, the unrealized gain now exceeds $20 billion — a return that would rank among the best venture-style investments ever made by a public technology company.

Beyond Intel: A Portfolio Strategy

The Intel investment, while the most visible, is one component of a broader strategy that Nvidia has been executing with increasing intensity. As we covered in our analysis of Nvidia’s full investment portfolio, the company is deploying capital across every layer of the AI value chain.

The pattern is consistent: identify companies whose growth depends on AI compute, take a significant equity position, and then deepen the commercial relationship through joint development, co-engineering, or platform integration. Each investment creates a financial incentive for the target company to standardize on Nvidia hardware and a financial return for Nvidia when the company succeeds.

This is not traditional venture capital. Nvidia is not a passive financial investor hoping for an IPO. It is a strategic investor using equity positions to build commercial moats — relationships that are reinforced by shared financial interests and become increasingly difficult for competitors to replicate.

The approach has also attracted regulatory scrutiny. The FTC’s review of the Intel transaction, while ultimately resulting in approval, signaled that antitrust authorities are paying attention to the dual role Nvidia now plays as both dominant chip supplier and major equity investor in the companies that buy those chips. Whether additional regulatory action follows will depend in part on how aggressively Nvidia conditions its commercial relationships on equity ties — a line that, so far, the company appears to have navigated carefully.

What It Means for the AI Industry

Nvidia’s investment empire has implications that extend well beyond its own balance sheet.

For AMD. Nvidia’s partnership with Intel on custom x86 CPUs and SOCs represents a competitive threat on multiple fronts. AMD has built its recent success on offering both CPUs and GPUs for data centers and PCs — a full-stack alternative to the Nvidia-plus-Intel combination that many servers traditionally used. If Nvidia and Intel deliver tightly integrated CPU-GPU products that outperform AMD’s offerings, the competitive dynamics in both data centers and client PCs could shift.

For the broader AI ecosystem. Nvidia’s equity investments create a gravitational pull that draws companies deeper into its ecosystem. Startups evaluating whether to build on Nvidia’s CUDA platform or explore alternatives now face an additional consideration: taking Nvidia investment means access to capital, technical resources, and commercial relationships, but it also means deeper dependence on a single hardware vendor. The AI industry’s structural reliance on Nvidia — already a subject of concern among some policymakers and technologists — is being reinforced through financial ties as well as technical ones.

For cloud providers. Amazon, Google, and Microsoft have all invested heavily in custom chip efforts designed to reduce their dependence on Nvidia GPUs. Nvidia’s investment strategy can be read in part as a response: by locking in equity relationships with key companies across the AI stack, Nvidia is building a network of aligned partners that can collectively resist the gravitational pull of the hyperscalers’ in-house silicon efforts. The Amazon-Anthropic investment is one example of how the major cloud players are pursuing a similar strategy on their own side of the equation.

Market Reaction

The broader market context has been favorable for AI-related stocks. The S&P 500 extended its winning streak to six consecutive weeks, with the benchmark index hovering near all-time highs above 7,300. Technology and semiconductor names have been among the primary beneficiaries of the rally, driven by strong earnings, easing trade tensions, and continued institutional inflows into AI-exposed equities.

In adjacent market action this week:

Monday.com surged 26% after reporting first-quarter results that beat analyst expectations across the board. Revenue grew 24% year-over-year to $351.3 million, with management crediting the company’s AI platform features for driving adoption among enterprise customers. The earnings beat and forward guidance prompted multiple analyst upgrades and added to the broader narrative that AI adoption is translating into measurable revenue growth across the software industry, not just for chip makers.

Lumentum announced it will join the Nasdaq 100 index on May 18, replacing CoStar Group. The inclusion reflects Lumentum’s rising market capitalization and the growing weight of optical networking and photonics companies in major indices — another signal that the AI infrastructure buildout is reshaping the composition of benchmark indexes. Index inclusion typically drives incremental buying from passive funds that track the Nasdaq 100, providing a technical tailwind for the stock.

Nvidia’s own shares have benefited from the combination of portfolio gains, the Intel partnership developments, and the broader market tailwind. The stock remains one of the most widely held positions among institutional investors, and the $40 billion investment portfolio adds a layer of asset value that is not always captured in traditional earnings-based valuation models.

What to Watch

Several catalysts in the coming weeks and months will determine whether Nvidia’s investment strategy continues to deliver:

Intel product details. The “exciting new products” teased by CEO Lip-Bu Tan need to be followed by concrete specifications, pricing, and availability timelines. If the NVIDIA-custom x86 CPUs and RTX-integrated SOCs prove competitive with AMD’s and Apple’s offerings, the Nvidia-Intel partnership could reshape the PC and server markets. If the products disappoint or face significant delays, the partnership narrative weakens.

Regulatory developments. The FTC approved the Intel equity acquisition, but the broader question of whether Nvidia’s dual role as chip supplier and ecosystem investor creates anticompetitive effects remains open. Congressional interest in AI market concentration has been growing, and Nvidia’s expanding portfolio could attract additional scrutiny — particularly if competitors file formal complaints.

OpenAI’s trajectory. Nvidia’s largest single bet, at roughly $30 billion, depends heavily on OpenAI’s ability to continue growing revenue and maintaining its position in the generative AI market. Competition from Anthropic, Google, and open-source alternatives has intensified. Any significant setback at OpenAI would impact the portfolio’s overall valuation.

Intel stock performance. The $5 billion stake that is now worth over $25 billion represents an extraordinary gain, but it is unrealized. Intel’s stock price will be influenced by the Apple foundry deal’s progress, the Nvidia product collaboration, and Intel’s own financial results. Any reversal in Intel shares would reduce the portfolio’s headline value.

Private portfolio markdowns or markups. The $17.5 billion deployed into private companies and infrastructure funds is valued based on the most recent funding rounds. As those companies seek additional capital or approach public listings, updated valuations could move in either direction.

Bottom Line

Nvidia has built something that no other semiconductor company has attempted at this scale: a $40 billion investment portfolio that doubles as a commercial strategy. By taking equity stakes in the companies that buy its chips, build its infrastructure, and develop applications on its platform, Nvidia has created financial and commercial alignment that reinforces its market dominance through channels that pure product competition cannot easily disrupt.

The Intel stake — bought at $5 billion and now worth five times that — is the most dramatic example, but it is the breadth of the strategy that matters. From OpenAI to Corning to dozens of private startups, Nvidia is embedding itself financially across every layer of the AI economy. Whether that produces long-term returns or concentrated risk depends on whether the AI buildout continues at its current pace. For now, the market is betting that it will — and Nvidia has positioned itself to profit from nearly every dollar spent in the process.

This article is for informational purposes only and does not constitute investment advice. Stock prices and portfolio valuations referenced are based on publicly available data as of May 12, 2026.