Mag 7 AI spending faces ROI reckoning as 88% boost budgets

A growing divide between hyperscaler and chipmaker stocks is forcing investors to confront whether the Mag 7's massive AI spending can deliver returns.

The Mag 7's AI spending spree is colliding with investor demands for returns, raising the risk of a divergence between hyperscalers and chipmakers that have rallied in tandem through the AI boom.

"Can both hyperscaler and chipmaker stocks rally together? That's the question investors need to answer as CapEx continues to climb," said Nathan Peterson, director of derivatives strategy at Charles Schwab.

The concern comes as 88% of organizations plan to increase AI investment, according to a November 2025 Gartner survey of 469 CEOs, while Gartner estimates AI coding costs will surpass the average developer's salary by 2028. Three-quarters of executives see tech budgets rising this year, with nearly half projecting double-digit jumps. Uber burned through its entire 2026 AI budget in just four months after employees rushed to adopt AI coding tools, forcing management to cap usage, according to reports.

The divergence risk is acute because chipmakers capture tangible revenue from AI infrastructure buildout while hyperscalers face mounting depreciation costs without clear monetization timelines. Memory chip prices have surged — Micron's DRAM prices rose more than 60% quarter-over-quarter and NAND flash prices increased more than 80% — benefiting manufacturers but squeezing AI developers and cloud providers. Apple this week raised prices on several MacBook and iPad models, citing rising memory costs. Tech stocks sold off for much of last week as investors reassessed AI valuations, with doubts about returns on massive spending compounded by a weak post-IPO showing from SpaceX and reports that OpenAI may delay its listing.

Chipmakers Gain as Hyperscalers Face Cost Squeeze

The divergence is already visible in market performance. Shares of Micron, SK Hynix, and Samsung have outperformed several major technology companies this year as demand for AI infrastructure accelerates. Micron said prices for its DRAM memory chips rose more than 60% in the quarter ended May 28 compared with the previous quarter, while NAND flash memory prices increased more than 80%. Shipment volumes rose only modestly, indicating that much of the revenue growth was driven by higher prices rather than volume expansion.

The global HBM market is dominated by three manufacturers — Micron, Samsung, and SK Hynix — and building additional production capacity requires significant investment and can take several years, limiting near-term supply growth. Cloud providers and AI model developers, including Microsoft, Amazon, Alphabet, and Meta Platforms, continue investing heavily in AI infrastructure despite the higher component costs. Additional memory capacity is expected to come online over the next several years, though current supply constraints are likely to persist until new manufacturing facilities begin production.

Cost Pressures Reshape AI Model Selection

The cost spike is pushing more businesses toward cheaper alternatives. OpenRouter, an AI marketplace, saw open-source tokens processed on its platform jump to 65% in June from 34% in January, according to a Citi note. The four most popular models on OpenRouter are all Chinese, with DeepSeek holding the top spot. Chinese models charge as little as 18 US cents per million tokens, against US$4 on average for top US models.

"Open-source models are showing that they are 90% as good at 10% of the price," said Val Bercovici, chief AI officer at WEKA, which helps companies run AI models faster and cheaper. "We don't need to spend the premium tokens on every level of effort."

The shift toward cheaper models could pressure revenue growth at OpenAI and Anthropic as they prepare for potential IPOs. "There will be a price-war dynamic when it comes to OpenAI and Anthropic as they both duke it out for a 'first to public market' IPO dates," said Christopher Brown, financial adviser at Synovus Securities, which owns shares in several Big Tech companies.

This article is for informational purposes only and does not constitute investment advice.