Nvidia's Falling Gross Margin Signals Pricing Power Shift to Memory Makers

Nvidia's latest quarterly results delivered staggering growth figures — $96.2 billion in total revenue, more than double the year-ago period — but the most consequential detail for the broader AI supply chain emerged in the company's forward guidance. Management expects gross margins to compress from 75% to between 71% and 72% by the fourth quarter, and executives explicitly attributed the decline to rising memory costs.

That margin pressure represents a fundamental shift in bargaining power within the AI infrastructure stack. As Nvidia scales production of its most advanced GPUs, the company is paying premium prices for the high-bandwidth memory modules that sit alongside those processors. This dynamic creates a direct pricing tailwind for memory manufacturers, particularly Micron Technology, which supplies the specialized DRAM and NAND products required for AI workloads.

Why it matters

Nvidia's margin guidance offers a rare window into cost structure shifts within the AI hardware ecosystem. When the dominant player in AI chips signals that a key input is commanding higher prices, it validates the investment thesis for upstream suppliers. For memory makers that have struggled with cyclical pricing pressure in recent years, Nvidia's acknowledgment that memory is driving margin compression confirms that AI demand is creating genuine pricing power — not just volume growth.

Data center segment now exceeds Nvidia's total 2023 revenue

The scale of Nvidia's data center business has reached a threshold that redefines the company's profile. The segment alone generated $89 billion in the quarter, surpassing what Nvidia produced across all product lines just one year earlier. Hyperscale customers — the cloud providers building out AI infrastructure at massive scale — contributed $48.7 billion, up 102% year over year.

Operating income reached $63.7 billion, up 124% from the prior year, while net income came in at $59.7 billion, or $2.46 per diluted share. The company's third-quarter revenue guidance of $108 billion (plus or minus 2%) assumes no compute sales to China, yet still represents sequential growth that few companies at Nvidia's scale can sustain.

Memory costs drive margin outlook lower

Nvidia's GAAP and non-GAAP gross margins both came in at 75% for the second quarter. For the third quarter, management guided both metrics to 74%, with expectations that margins will bottom between 71% and 72% in the fourth quarter before recovering in fiscal 2028. The company's CFO attributed the decline specifically to memory pricing, according to details first reported by AI Watch.

This guidance matters because it quantifies the cost pressure Nvidia faces as it sources high-bandwidth memory for its Hopper and Blackwell architectures. The premium Nvidia is paying reflects tight supply conditions for advanced memory products, which require specialized manufacturing processes and capacity that cannot be added quickly. For investors evaluating memory suppliers, Nvidia's margin compression serves as confirmation that pricing leverage has shifted.

The details of Nvidia's fiscal 2027 second-quarter results and forward guidance were first reported by AI Watch.