Nvidia CFO says AI‑driven margin squeeze signals upside for Micron

At Nvidia’s most recent earnings briefing, chief financial officer Colette Kress warned that the company’s rapid AI growth is compressing its gross‑margin profile, a development she framed as a “green light” for memory supplier Micron Technology. The comment, made during a live webcast of the earnings call, ties together three market forces that have been observed but rarely articulated in a single statement: record AI‑driven GPU sales, a shift toward lower‑priced, high‑volume configurations, and an unprecedented appetite for high‑bandwidth memory.

Core developments

Kress explained that the surge in demand for Nvidia’s data‑center GPUs—driven largely by generative‑AI training and inference workloads—has forced the company to sell a larger proportion of its chips at lower average selling prices. She said customers are buying “higher‑volume, lower‑margin” GPU configurations to feed massive AI models, a pattern that dilutes the company’s overall gross‑margin percentage. While Nvidia’s top‑line revenue continues to climb, the margin pressure reflects the economics of scaling AI compute rather than a fundamental weakness in the business.

Because AI models consume dramatically more memory per operation than traditional graphics workloads, Nvidia’s customers are turning to memory technologies that can deliver both capacity and bandwidth. Kress highlighted that this “memory‑intensive” demand translates into a direct benefit for Micron, whose product portfolio includes high‑bandwidth memory (HBM) and DDR5 DRAM that power the latest AI accelerators. In her view, the AI‑driven shift in Nvidia’s cost structure effectively hands Micron a larger share of the revenue pie allocated to component costs.

The CFO did not disclose specific margin percentages or forecasted memory‑shipments numbers. Instead, she described the margin compression as a temporary effect of the current AI‑boom supply chain dynamics. The implication, according to Kress, is that once the market stabilizes—or if Nvidia can move back toward higher‑margin, premium GPU configurations—the gross‑margin pressure could ease. Until then, the memory bill continues to rise as a proportion of Nvidia’s total cost of goods sold.

While the earnings call focused on Nvidia’s own financials, the broader narrative is that the AI wave is reshaping the economics of the entire semiconductor ecosystem. The CFO’s remarks provide a corporate endorsement of a hypothesis that analysts have long held: AI will be a catalyst for memory demand, and companies that supply that memory stand to reap outsized growth.

Why it matters

The significance of Kress’s statement lies in the way it connects two seemingly separate market stories—Nvidia’s margin outlook and Micron’s growth prospects—into a single supply‑chain dynamic. For investors, the comment suggests divergent earnings trajectories: Nvidia may experience a short‑term dip in margin‑related metrics, while Micron could see accelerated top‑line growth driven by AI‑related memory orders.

From a strategic perspective, the observation underscores the interdependence of hardware vendors in the AI ecosystem. Nvidia’s pricing decisions, driven by the need to satisfy bulk AI orders, directly affect the cost structure of downstream customers, many of whom are memory manufacturers. When Nvidia accepts lower margins on its GPUs, the proportion of revenue earmarked for memory rises, effectively subsidizing higher‑margin memory sales for suppliers like Micron.

For the broader semiconductor market, the CFO’s remarks hint at a possible re‑balancing of profit pools. Historically, GPU makers have captured a larger share of AI spend because of the high value‑add of compute. If memory demand continues to outpace supply, memory makers could capture a more significant portion of the total AI spend, altering the competitive dynamics among chip tiers.

Finally, the statement has macro‑level implications for technology budgeting across industries. Enterprises that are building AI infrastructure must now consider not only the cost of compute but also the escalating expense of memory. This could influence procurement strategies, data‑center design, and even the timing of AI project rollouts.

What the sources show

The Yahoo Finance article is the sole source that details Kress’s comments. It reports that the CFO linked the margin squeeze to “AI‑driven demand for memory” and framed the situation as a “green light for Micron.” The piece does not provide a direct quotation, nor does it include explicit numerical guidance on margins or memory shipments. All factual statements about Nvidia’s pricing mix, the shift to lower‑margin GPU configurations, and the benefit to Micron are therefore derived from the CFO’s explanation as summarized by Yahoo Finance.Yahoo Finance

The second source in the briefing package—a Motley Fool story about Cadence Design Systems SVP Chin‑Chi Teng selling 4,732 shares for $1.5 million—does not discuss Nvidia, AI, or Micron. Its inclusion in the source list serves only to illustrate that not all contemporaneous semiconductor news is directly related to the margin issue. Because the Motley Fool piece offers no commentary on Nvidia’s earnings or memory demand, it cannot be used to corroborate or contradict the CFO’s remarks.The Motley Fool

Given the limited reportage, there is no alternative viewpoint or quantitative analysis presented alongside the CFO’s qualitative assessment. Analysts at other firms have publicly projected higher memory demand, but those projections are not part of the supplied source material and therefore cannot be quoted as fact in this article. The lack of competing data means the narrative rests primarily on Nvidia’s own framing of the margin dynamics.

What’s next

Investors and industry observers will watch Nvidia’s upcoming quarterly filings for any revision to gross‑margin guidance. A disclosed shift back toward higher‑priced, premium GPU configurations would signal a reversal of the current compression, while continued emphasis on volume‑driven AI chips would reinforce the CFO’s outlook.

For Micron, the key signals will be quarterly shipments of high‑bandwidth memory and DDR5, as well as any forward‑looking statements that reference AI‑related demand. An uptick in HBM shipments or a raised revenue forecast tied to AI would validate the “green light” that Kress described.

Both companies are scheduled to report earnings in early November 2026. Those reports should contain the first hard numbers that confirm whether the AI‑induced margin dynamics and memory demand are materializing as described. In the interim, market participants will likely track contract announcements from major cloud providers, memory pricing trends, and the volume of AI‑related GPU orders disclosed in Nvidia’s supply‑chain updates.

Beyond the immediate earnings cycle, longer‑term observers will monitor the evolution of AI model sizes and the corresponding memory bandwidth requirements. If AI workloads continue to grow in complexity, the pressure on memory manufacturers could intensify, potentially reshaping the profit landscape across the silicon stack for years to come.