Nvidia, Micron Post Blowout AI Earnings as Yields Roil Market

Nvidia and Micron's earnings arrive at a critical juncture for AI-focused investors. The 10-year Treasury yield sitting in the 92.8th percentile of the past year has historically pressured expensive growth stocks, yet both companies delivered results that reinforce their positions in the AI buildout. The body frames them as complementary: Nvidia controls the compute layer—Jensen Huang highlighted Microsoft's Fairwater site running hundreds of thousands of Blackwell GPUs and guided to $91 billion in Q2 revenue—while Micron holds the memory bottleneck that those systems cannot bypass.

Micron's structural innovation is the attempt to defang memory's historical cyclicality. The 16 Strategic Customer Agreements with roughly $100 billion in minimum contracted revenue and price floors above prior peak margins represent a departure from memory's boom-bust playbook. Customer cash deposits of $22 billion further lock in runway. For comparison, Nvidia's $145 billion in supply commitments and $1 trillion revenue forecast through calendar 2027 emphasize platform economics, though Huang noted China Data Center compute exclusions from guidance—a real overhead for the stock.

The market has already begun to parse this distinction. Nvidia fell 3.75% last week while Micron gained 2.58% and is up 241.59% year-to-date, reflecting the yield shock hitting high-multiple growth names harder. The body positions the two as responses to different investor profiles: Micron for value-focused investors seeking compressed multiples and rare memory-market visibility, Nvidia for those seeking the purest AI compounder willing to tolerate multiple compression during yield spikes.