Nvidia CEO Jensen Huang recently spoke at the Morgan Stanley Technology, Media and Telecom Conference, discussing his company's position in AI and data centre markets. His message was striking: in a world of constraint, scarcity is an advantage.
The company's AI hardware is raking in money, driving electronics supply chain pressures including memory shortages and wafer production capacity strain. Nvidia's AI factory revolution is straining current supply lines to their limits. Yet rather than apologising for the crunch, Huang framed it as opportunity. In a constrained world, he argued, customers have no choice but to pick the best. That usually means Nvidia.
This confidence may look tone-deaf, but it's backed by concrete action. Nvidia is investing a combined $4 billion in photonics companies Coherent and Lumentum, with each company receiving $2 billion. The deals underscore how the company plans to stay ahead as supply chains tighten and competition heats up.
Securing the Pipeline
The investments help Nvidia bolster its data centre chips with photonics technology that uses light rather than electrical signals for connections between AI chips. This matters because both companies develop optical and semiconductor technologies, enabling future networking hardware and AI chip designs that utilise light to transfer data, rather than electricity.
The Lumentum deal includes a multiyear strategic agreement with multibillion-dollar purchase commitments and future capacity rights for advanced laser components. With Coherent, Nvidia is working on next-generation silicon photonics for AI infrastructure, also including multibillion-dollar purchase commitments and future access rights.
The $2 billion Lumentum investment will support R&D and manufacturing expansion with a new US fabrication facility, plus multibillion purchase commitments and future capacity access for advanced lasers. The Coherent deal includes $2 billion for US-based manufacturing expansion.
Controlling Supply, Not Just Demand
The strategy reveals something important about Huang's scarcity comment. Nvidia isn't just benefiting from shortage; it's working to prevent one from choking off its own supply chain. Nvidia's investment in expanding photonics development and fabrication in America may be a way to get ahead of future supply constraints.
This isn't exclusivity. Nvidia secured capacity but not exclusive rights. It isn't looking to eat the AI industry and supply chain, but bolster its output. Both Lumentum and Coherent will continue serving other customers. But Nvidia gets priority access and a seat at the table shaping the next generation of optical AI infrastructure.
The timing matters. Meta, one of Nvidia's top customers, signed a $60 billion deal with rival Advanced Micro Devices last week, fuelling the need to improve performance of semiconductors. Nvidia's investments in photonics help it stay ahead of AMD and other competitors chasing similar deals.
The Pragmatic Middle Ground
Huang's comments about scarcity benefiting dominant players contain an uncomfortable truth. When supply can't meet demand, price discipline erodes and the leader gains bargaining power. That's good news for Nvidia's shareholders but less encouraging for customers and the broader industry paying inflated prices for constrained capacity.
Yet Nvidia's approach here suggests a more measured view. Rather than hoarding supply or forcing exclusive arrangements, the company is investing in expanding capacity itself. Nvidia has been working with both companies for some time on networking hardware and supply chain development; these new deals only enhance that cooperation. The multibillion-dollar commitments signal long-term intent, not short-term speculation.
That said, the investments also protect Nvidia's market position. By investing heavily in photonics now, it shapes industry standards around light-based interconnects at a time when other players are moving in. China has used Nvidia's absence to fuel domestic inference chips. AMD recently signed $60 billion with Meta, and startups are building custom AI chips.
The real question is whether supply will remain genuinely constrained or whether Nvidia's investments help break the bottleneck. If photonics capacity expands significantly, scarcity becomes less of an advantage. Huang is betting he can stay ahead even if the supply crunch eases, by building the best infrastructure and locking in long-term customer relationships. That's a more durable strategy than simply profiting from shortage while it lasts.