Inside Nvidia’s $18.4 Billion Public Stock Portfolio: Should Investors Follow?

Nvidia’s latest 13F filing is not just a list of stocks. It is almost a map of where the company thinks the next bottlenecks in the AI boom will appear. In three months, Nvidia’s disclosed public stock portfolio jumped from $13.1 billion to $18.4 billion, expanded from 5 names to 7 names, and quietly revealed a bigger idea: Nvidia is not only selling the picks and shovels of AI. It is also investing in the roads, power lines, design tools, cloud platforms and network pipes that can make demand for those picks and shovels last longer.

Let's break down where Nvidia is investing, why these companies matter, what changed between the February and May filings, how these stocks have moved since the latest disclosed holding date, and whether investors should treat Nvidia’s portfolio as a buy list or as something more useful: a signal map.

What Is Nvidia’s 13F Portfolio?

A 13F is a quarterly filing that shows certain US-listed equity holdings of large institutional managers. But investors should be careful here. A 13F is not real-time. It shows holdings as of the quarter-end and is usually filed weeks later. It also does not show everything a company owns. Private investments, some preferred stock deals, debt instruments and positions after the quarter-end may not appear in the filing.

So, Nvidia’s May 2026 13F tells us what it held as of March 31, 2026. It does not tell us what it bought or sold yesterday. Think of it like looking at a cricket scorecard after the innings is over. You can understand the strategy, but you cannot assume the next ball will be the same.

That said, Nvidia’s filing still matters because this is not a normal financial portfolio. Nvidia is not Warren Buffett looking for undervalued consumer stocks. Nvidia is the most important AI infrastructure company in the world, and its equity bets often sit close to its commercial strategy.

Nvidia’s Latest 13F Holdings: 7 Stocks In Its Portfolio

The May filing shows 7 disclosed holdings worth $18.37 billion. The February filing had 5 holdings worth $13.10 billion. The biggest change is simple: Nvidia added Coherent and Generate Biomedicines, and nearly doubled its CoreWeave share count.

The table below uses Nvidia’s March 31, 2026 13F value and share count. The return column compares the 13F-implied March 31 price with the latest available June 1, 2026 market price. For Generate Biomedicines, the current value estimate is based on available ownership tracking data.

Source: Nvidia’s 13-F Filing

The first big takeaway: Nvidia’s portfolio is extremely concentrated. Intel alone is more than half of the disclosed portfolio. CoreWeave, Synopsys, Coherent and Nokia together make up almost the entire remaining portfolio. Generate Biomedicines is tiny in dollar terms, but strategically interesting because it shows Nvidia’s AI thesis moving beyond data centres into scientific discovery.

Why Nvidia Is Investing In AI Infrastructure Bottlenecks

The easiest way to misunderstand this portfolio is to call it an “AI stock basket.” It is not. A better way to read it is as a bottleneck portfolio.

Every major AI boom creates constraints. First, everyone needed GPUs. Then they needed power. Then data centres. Then networking. Then memory. Then software tools to design faster chips. Then cloud providers to rent compute. Nvidia’s public portfolio seems to be pointing toward the places where those constraints may show up next.

Think of Nvidia like a company building India’s largest expressway. If traffic is going to multiply, the builder does not only care about the road. It also cares about toll booths, power substations, petrol pumps, traffic control systems and logistics hubs. That is what this portfolio resembles. Nvidia is investing around the AI expressway.

Nvidia Portfolio Stocks: Why Each Holding Matters

Intel (INTC): The Strategic Anchor

Intel is the largest position by far. Nvidia’s partnership with Intel is built around custom data centre CPUs and PC products that combine Intel’s x86 strength with Nvidia’s AI and GPU ecosystem. For Nvidia, this is not simply a bet on Intel’s turnaround. It is a bet on tighter CPU-GPU integration.

Why does that matter? AI systems are not only about GPUs. CPUs still coordinate workloads, move data, manage systems and support enterprise software. If Nvidia can connect its AI platforms more deeply with Intel’s x86 base, it can reach more data centre and PC customers without forcing the world to abandon the architecture it already uses.

The risk is that Intel remains a difficult turnaround story. Manufacturing delays, foundry pressure, competition from AMD and Arm-based CPUs, and execution gaps can still hurt the stock. But from Nvidia’s point of view, Intel gives strategic reach that a normal portfolio manager may not fully capture.

CoreWeave (CRWV): The AI Factory Accelerator

CoreWeave is Nvidia’s clearest “demand creation” investment. CoreWeave builds AI cloud infrastructure around Nvidia GPUs. Nvidia nearly doubled its share count from the February filing to the May filing, making this the most important active increase in the portfolio.

The logic is straightforward. If AI companies need more compute, someone has to build the cloud capacity. Hyperscalers like Microsoft, Amazon and Google are doing it, but specialised AI cloud companies like CoreWeave can move faster in certain workloads. For Nvidia, CoreWeave helps convert GPU demand into deployed AI factories.

An AI factory is basically a data centre that turns electricity, chips and software into AI output. Just like a textile mill turns cotton into fabric, an AI factory turns compute into tokens, models and inference. The more these factories scale, the bigger Nvidia’s long-term market can become.

But CoreWeave is also one of the riskiest names. It is capital heavy, debt heavy, and exposed to customer concentration. If AI demand slows or customers build more in-house capacity, CoreWeave can face pressure quickly.

Synopsys (SNPS): The Toolmaker Behind The Toolmaker

Synopsys may look less exciting than CoreWeave, but it is one of the most important companies in the semiconductor world. It provides electronic design automation software and design IP used by chip companies to build complex semiconductors.

Nvidia’s investment in Synopsys is about accelerating engineering and design. As chips become more complex, the software used to design them becomes more valuable. In simple terms, if Nvidia is building faster cars, Synopsys helps build better factories and simulation systems for those cars.

This is a lower-drama strategic bet. It does not directly create GPU demand like CoreWeave, but it can improve the pace of chip design, digital twins, simulation and physical AI applications. The risk is valuation and execution, especially after Synopsys’ large Ansys acquisition. But strategically, Synopsys sits close to the heart of the AI hardware cycle.

Coherent (COHR): Solving The Data Movement Problem

Coherent is one of the two new positions in Nvidia’s May filing, and it may be the most important new signal. Coherent works in lasers, optical components and photonics. In AI data centres, moving data fast and efficiently is becoming just as important as computing it.

This is where silicon photonics matters. It uses light to move data, helping data centres handle huge AI workloads with better speed and energy efficiency. A simple way to understand it: if GPUs are powerful factories, optical networking is the highway connecting those factories. A jammed highway can slow the entire economy.

Nvidia’s Coherent investment suggests the company is thinking seriously about the networking layer inside next-generation AI data centres. The risk is that optics can be cyclical, competitive and technically demanding. But if AI clusters keep getting larger, faster data movement becomes a structural need, not a side issue.

Nokia (NOK): The Edge AI Bet

Nokia is not an obvious Nvidia holding until you look at AI-RAN and 6G. AI-RAN means artificial intelligence built into radio access networks, the part of telecom infrastructure that connects phones and devices to the network. Nvidia and Nokia are working on AI-native networks, with Nvidia investing in Nokia to support the transition from 5G to 6G.

Why does this matter? Today, most AI compute sits in data centres. But over time, AI will also move closer to devices, robots, cars, factories and mobile networks. That requires edge infrastructure. Nokia gives Nvidia exposure to that future.

For Indian investors, think of this like the jump from 3G to 4G, and then to 5G. Each network upgrade did not just improve mobile speed. It created new business models. 6G and AI-RAN could do something similar for machine intelligence, although the timing is still uncertain.

Nebius (NBIS): The Smaller Cloud Signal

Nebius is a much smaller 13F position, but strategically it fits the same pattern as CoreWeave. It is a full-stack AI cloud company trying to build large-scale AI infrastructure. Nvidia’s partnership with Nebius focuses on AI factories, inference and agentic AI workloads.

The interesting part is that Nvidia is backing multiple AI cloud players, not just one. That reduces dependence on hyperscalers and creates more paths for Nvidia systems to reach customers. But Nebius also carries similar risks: huge capital needs, execution pressure and sensitivity to AI demand cycles.

Generate Biomedicines (GENB): The Option Bet

Generate Biomedicines is tiny in Nvidia’s portfolio, but it adds an important message. Nvidia is not only investing in the infrastructure of AI. It is also putting small chips on industries where AI could create entirely new demand.

Generate uses machine learning to design and optimise proteins for medicines. This is a very different type of bet from Intel or CoreWeave. It does not directly build data centres or networking. Instead, it sits in the “what will people do with all this compute?” bucket.

Drug discovery is one of the most compute-hungry scientific fields. If AI helps speed up molecule or protein design, demand for accelerated computing can expand from internet companies to healthcare, pharma and research labs. The risk is obvious: biotech outcomes are uncertain, clinical trials can fail, and stock moves can be brutal.

What Changed In Nvidia’s Latest 13F Filing?

The biggest change is that Nvidia moved from a 5-stock portfolio to a 7-stock portfolio. Coherent and Generate Biomedicines were new. CoreWeave was almost doubled. Intel, Synopsys, Nokia and Nebius share counts stayed unchanged.

That tells us three things.

• First, Nvidia is still extremely confident in CoreWeave’s AI cloud buildout. The near-doubling of shares is not a small signal.
• Second, Nvidia is expanding from compute into connectivity. Coherent and Nokia show that Nvidia sees networking, optics and telecom as part of the AI infrastructure story.
• Third, Nvidia is keeping optionality in AI applications. Generate is small, but it hints at where future compute demand may come from: biology, science and industrial research.

Should Investors Follow Nvidia’s AI Stock Bets?

Investors should follow Nvidia’s thesis, but not blindly follow its trades.

That distinction matters. Nvidia can benefit from these companies in ways a retail investor cannot. If CoreWeave buys more Nvidia GPUs, Nvidia benefits commercially even if CoreWeave’s stock becomes volatile. If Intel builds custom CPUs for Nvidia platforms, Nvidia gains strategic flexibility even if Intel’s turnaround takes longer. If Nokia helps create AI-RAN demand, Nvidia may benefit through platform sales, not only through Nokia’s stock price.

Retail investors do not get those side benefits. They only get the stock’s risk and return.

So the better approach is to use Nvidia’s 13F as a research shortlist. Ask three questions before considering any name:

• Does this company remove a real bottleneck in the AI value chain?
• Does Nvidia’s partnership improve the company’s business fundamentals, not just its market narrative?
• Can the company stand on its own if Nvidia’s halo effect fades?

If the answer is yes, the stock may deserve deeper study. If the answer is only “Nvidia owns it,” that is not enough.

Key Risks In Nvidia’s AI Infrastructure Investment Thesis

• The biggest risk is that AI infrastructure spending slows before these companies earn enough return on capital. Many AI cloud and data centre businesses need massive upfront investment. If customers reduce orders, delay projects or shift to cheaper in-house chips, the economics can change fast.
• The second risk is circularity. Nvidia invests in companies that may use Nvidia technology, which can make investors question whether demand is fully organic. Nvidia argues these investments are strategic and small relative to the capital required, but the market will keep watching this closely.
• The third risk is that bottlenecks move. Today, optics, power, cloud and networking look scarce. Tomorrow, the scarce layer may be memory, custom silicon, software efficiency or energy contracts. A portfolio built around today’s bottlenecks can underperform if the bottleneck shifts.

Bottom Line: Nvidia’s 13F Portfolio Is A Signal, Not A Buy List

Nvidia’s public stock portfolio is not a simple copy-trading list. It is a window into how the world’s most important AI infrastructure company sees the next stage of the AI cycle.

The message is clear: Nvidia believes the AI boom is moving from chips alone to full-stack infrastructure. That includes cloud capacity through CoreWeave and Nebius, CPU integration through Intel, chip design through Synopsys, optics through Coherent, edge networks through Nokia, and scientific AI applications through Generate Biomedicines.

For investors, the smartest takeaway is not “buy everything Nvidia owns.” The smarter takeaway is this: study the bottlenecks Nvidia is funding. In every major technology cycle, the biggest returns often come not only from the headline winner, but from the companies that quietly make the winner’s growth possible.

Nvidia’s 13F is useful because it shows us where those enabling layers may be forming. But the final investment decision still has to come from valuation, balance sheet strength, execution, and your own risk appetite.

In short, Nvidia’s portfolio is worth following as a map. It is not worth copying as a shortcut.