Nvidia's latest SEC filing reveals the company holds a roughly $21 billion stake in SpaceX and a $30 billion stake in Intel — positions that, combined, dwarf the market capitalizations of most semiconductor firms2,4. ANALYSIS The disclosure marks a structural shift: the dominant supplier of AI training chips is now deploying GPU-era cash as strategic capital across the AI stack, blurring the line between chipmaker and conglomerate.
Why it matters
Nvidia's core business sells the picks and shovels of the AI boom. ◆ But the filing shows the company is no longer content to collect revenue from chip sales alone — it is taking equity positions in the very companies that consume its hardware, creating financial entanglements that could reshape competitive dynamics across AI infrastructure. Separately, Lam Research's plan to invest more than $3 billion over five years to expand its global R&D network signals that capital is flooding not just into chips but into the equipment that makes them1,3.
The big picture
Nvidia disclosed in a filing with the Securities and Exchange Commission that it owns 122.8 million Class A shares in SpaceX. That stake originated from Nvidia's $10 billion investment in xAI as part of a $20 billion round in January. SpaceX acquired xAI in February in a deal valued at $1.25 trillion, converting Nvidia's xAI position into SpaceX equity. SpaceX held its public market debut in June.
The SpaceX position is Nvidia's second-largest holding behind Intel. Nvidia invested $5 billion in Intel, and that stake was worth about $30 billion at the end of the second quarter. Nvidia's Intel stake is currently worth about $22 billion. Nvidia also invested as much as $2 billion in xAI in 2025.
Meanwhile, the relationship between Nvidia and SpaceX is tightening on the product side. Elon Musk said SpaceX will exclusively use Nvidia chips in its AI data centers and to power its AI frontier models. Musk expects SpaceX will receive a significant allocation of Nvidia's Vera Rubin GPUs next year.
ANALYSIS The pattern is striking: Nvidia invested cash into xAI, that position converted into SpaceX equity through the acquisition, and SpaceX then committed to exclusive use of Nvidia hardware. The capital and the product relationship now reinforce each other.
On the equipment side of the semiconductor supply chain, Lam Research plans to spend more than $3 billion over the next five years to enhance its engineering capabilities. The company aims to increase its experiment capacity by more than 50%. Lam sells machines that chipmakers use to produce memory chips, advanced packaging, and other products.
Between the lines
Nvidia invested $5 billion in Intel, and its $10 billion xAI investment converted into SpaceX shares after the acquisition. ANALYSIS These are not passive financial plays. Nvidia's Intel stake sits alongside a company that fabricates semiconductors, while the SpaceX stake connects Nvidia to a compute-infrastructure operation that has committed to running exclusively on Nvidia hardware — each position anchoring a different node in the AI supply chain.
SpaceX's stock closed at $140 on Friday, down from $170.86 at the end of June, so the value of Nvidia's shares has declined to about $17.2 billion. ◆ The volatility underscores a new kind of risk for Nvidia: its financial results now carry mark-to-market exposure to companies it cannot control, a dynamic more familiar to holding companies than to hardware vendors.
Lam Research's $3 billion R&D expansion fits the broader picture. ◆ As chipmakers race to build AI-optimized silicon, the upstream equipment makers are scaling capacity to meet demand — a second-order capital cycle that mirrors Nvidia's own downstream investment activity. Capital is being deployed at every layer of the stack simultaneously.
What's next
Musk's expectation of a significant allocation of Nvidia's Vera Rubin GPUs next year will test whether Nvidia can serve a customer in which it holds a multi-billion-dollar equity stake without raising concerns from other large buyers. ◆ Nvidia's dual role — as supplier and as shareholder — will face increasing scrutiny as these positions grow. Lam Research's five-year investment timeline means the equipment buildout will unfold alongside Nvidia's own product roadmap, linking the fortunes of chip designers, fabricators, and toolmakers more tightly than at any point in the industry's history.