Nvidia and Tesla are reportedly planning to invest 3.7 trillion won (about $2.8 billion) in Samsung Electro-Mechanics to expand production of advanced semiconductor substrates—a small but essential component inside AI hardware. The report, first published by South Korea's Maeil Business Newspaper and picked up by Reuters, suggests the money would fund new substrate lines at Samsung Electro-Mechanics' facilities in Sejong and Busan, under a multi-year supply agreement tied to Nvidia.
Substrates are the foundation that connects and supports cutting-edge chips, acting as a bridge between the silicon die and the rest of the circuit board. They are critical in Nvidia's AI processors and high-speed networking switches, and they are notoriously difficult to scale up quickly because improving output requires specialized equipment and time. This is why big tech companies are increasingly willing to put money directly into suppliers—to ensure they get the components they need when they need them.
What the deal would look like
According to the report, Samsung Electro-Mechanics expects shipments to Nvidia to begin in 2026, with volumes growing at a double-digit annual pace. By 2028, the Nvidia business could take up a meaningful share of the company's substrate capacity. Tesla has not confirmed the report, but the broader message is clear: major buyers are moving upstream, using funding and long-term contracts to lock in the “unsexy” components that can slow an entire AI buildout.
For Nvidia, this kind of buyer-backed expansion reduces the risk that a parts shortage delays shipments once the new lines are running. For Samsung Electro-Mechanics, a dedicated-line deal can improve revenue visibility—knowing that a large customer is committed for years—but it also raises customer-concentration risk. If a larger slice of future output is earmarked for one client, swings in that client's ordering plans matter more.
Why this matters for investors
The report highlights a shift investors are watching across semiconductors: competitive advantage is increasingly about supply assurance, not just chip design. As AI demand surges, the companies that can guarantee access to critical components—whether substrates, memory, or advanced packaging—may have an edge over those that rely on spot markets.
For everyday investors, this is a reminder that the AI boom isn't just about the chip designers themselves. The supply chain behind them—companies making substrates, packaging, and other specialized parts—can be just as important. When big tech firms invest directly in these suppliers, it can signal confidence in future demand and provide a more stable revenue stream for the supplier.
However, it also means that these suppliers become more exposed to the fortunes of a single customer. If Nvidia's orders slow down, Samsung Electro-Mechanics could feel the impact more acutely. Investors should watch how much of the company's capacity is tied to Nvidia and whether other customers are also lining up.
The reported investment comes as Tesla's own AI and robotaxi bets are driving up costs, and as the AI boom reshapes the tech landscape. It also follows other big-ticket supply-chain moves, like LSEG's acquisition of data firm Octus, showing that companies are willing to spend heavily to secure what they need.
What to watch next
Investors will be watching for official confirmation from Samsung Electro-Mechanics, Nvidia, or Tesla. The report is based on unnamed sources, so the details could change. If confirmed, the deal would be a significant vote of confidence in the substrate supply chain and could prompt other chipmakers to consider similar arrangements.
For now, the key takeaway is that the AI supply chain is becoming a battleground. Companies are no longer just competing on design; they are competing on access to the components that make AI possible. This deal, if it goes through, would be a clear example of that trend.
