Nvidia Commits $279 Billion in Supply Chain Investments Focused on Memory Chips
During its second-quarter earnings report, Nvidia revealed plans to spend up to $279 billion over the next few years on supply commitments—more than double from just three months ago. The increase is primarily driven by demand for high bandwidth memory (HBM), a critical component in AI infrastructure that remains scarce as generative models grow more complex. Nvidia's data center segment, which saw $89 billion in revenue during the second quarter of fiscal 2027—a year-over-year jump of 117%—is central to this strategy. The company is pre-paying for memory chips over a multi-year period to ensure supply stability and support its Blackwell systems and Vera CPUs. This move signals strong confidence in AI growth but also indicates potential margin pressure due to rising HBM costs. The bulk of Nvidia's spending will likely go toward SK Hynix and Micron Technology, the two main suppliers qualified for HBM used in its platforms. The company expects gross margins to decline slightly over time as a result of higher memory prices.
