The AI supply chain is crunched not just by obvious components like TSMC wafers and HBM memory. A significant, often overlooked bottleneck is rack manufacturing—including high-speed cables, connectors, and even sheet metal—which are "sneaky hard" due to extreme power, heat, and signal integrity demands.

Hyperscalers are selling their own securities (stocks, bonds) to fund a massive CapEx cycle in physical infrastructure. The most direct trade is to mirror their actions: sell their securities and buy what they are buying—the raw materials and commodities needed for data centers, where the real bottlenecks now lie.

Companies like Tesla and AWS are investing in lithium and copper refining to control their supply chains, a new phase of vertical integration driven by AI's massive industrial needs for data centers and batteries.

While the world focused on GPU shortages, the real constraint on AI compute is now physical infrastructure. The bottleneck has moved to accessing power, building data centers, and finding specialized labor like electricians and acquiring basic materials like structural steel. Merely acquiring chips is no longer enough to scale.

While data was once a major constraint for training AI, models can now effectively create their own synthetic data. This has shifted the critical choke points in the AI supply chain to physical infrastructure like power grids and data center construction, which are now the primary limiters of growth.

The AI boom has created a series of supply chain bottlenecks. First, it was GPUs (Nvidia), then energy (GE Vernova), and now fiber optic cables (Corning). Companies that solve these critical shortages command immense pricing power, leading to soaring stock prices. The key is to find the next essential, scarce component.

The current commodity supercycle is intensified because traditionally asset-light tech companies (hyperscalers) are now massive consumers of physical resources. They are building data centers and competing for materials like copper, fundamentally altering their business models and commodity demand.

The race to build AI infrastructure was constrained not by advanced semiconductors, but by the availability of power transformers. This overlooked, 100-year-old technology saw lead times balloon to over three years, becoming the single biggest gating factor for new data center deployments.

The primary constraint on the AI boom is not chips or capital, but aging physical infrastructure. In Santa Clara, NVIDIA's hometown, fully constructed data centers are sitting empty for years simply because the local utility cannot supply enough electricity. This highlights how the pace of AI development is ultimately tethered to the physical world's limitations.