Nvidia and Alphabet are sitting out the megacap tech rally as mounting anxiety over AI infrastructure spending triggers a sharp selloff in semiconductor stocks. While the broader tech sector claws back losses, chip makers are getting hammered – with SoftBank Group leading a steep decline across Asian tech markets. The divergence signals a dramatic shift in investor sentiment about who’s actually profiting from the AI boom versus who’s just footing the bill.

The megacap tech trade just split wide open. Nvidia and Alphabet are getting crushed while their peers mount a comeback, and the culprit is a sudden crisis of confidence around AI infrastructure costs.

SoftBank Group led the carnage in Asian markets, dragging down a swath of chip and AI-adjacent stocks as investors question whether the massive capital outlays required to build AI infrastructure will ever generate returns that justify the spending. It’s a stark reversal from the AI euphoria that’s dominated tech markets for the past year.

The divergence is telling. While companies like Apple and Microsoft – which are integrating AI into existing products rather than building the underlying infrastructure – are bouncing back, the picks-and-shovels plays are getting hammered. Nvidia, the poster child of the AI chip boom, is bearing the brunt as investors recalculate the trajectory of data center spending.

Alphabet finds itself caught in the crossfire. The company’s been pouring billions into AI infrastructure through Google Cloud while simultaneously racing to integrate AI across its product suite. Now investors are asking hard questions about whether those investments will pay off – especially as competition from Microsoft and OpenAI intensifies.

The Asian selloff spreading from SoftBank reflects deeper concerns. SoftBank’s portfolio is heavily weighted toward AI infrastructure plays through its Vision Fund, and the company’s stock performance often serves as a proxy for broader sentiment around tech investment cycles. When SoftBank plunges, it’s usually signaling that investors are getting nervous about valuations in the entire ecosystem.

What’s driving the sudden shift? Mounting evidence that AI infrastructure costs are spiraling faster than revenue growth. Cloud providers are spending tens of billions on Nvidia chips, power infrastructure, and data centers, but actual AI product revenue remains relatively small. The math is starting to look uncomfortable.

Chip stocks are particularly vulnerable because they sit at the beginning of the value chain. If companies start pulling back on AI infrastructure spending – or even just slowing the rate of growth – chip makers feel it first. Nvidia has enjoyed monster growth as the dominant supplier of AI training chips, but that dominance also means the company has the most to lose if spending patterns shift.

The megacap divergence also reflects a broader debate about where value accrues in the AI stack. Are chip makers and infrastructure providers capturing sustainable margins, or are they just facilitating a gold rush where the real winners will be application-layer companies? Right now, markets are betting on the latter.

Microsoft and Apple represent the other side of that trade. Both companies are integrating AI into massive existing user bases – Microsoft through Copilot and Azure, Apple through on-device AI features. They’re spending on infrastructure too, but they have clear paths to monetization through subscriptions and device sales. That’s looking a lot more attractive than pure infrastructure plays right now.

The timing is particularly awkward for Alphabet, which just finished reassuring investors about its AI strategy after stumbling out of the gate with Bard. The company’s been making real progress with Gemini and catching up to OpenAI, but now it faces questions about whether all that catch-up spending was worth it.

Asian markets are amplifying the concerns because the region is home to critical parts of the AI supply chain. When SoftBank leads a selloff, it often pulls down semiconductor equipment makers, chip designers, and manufacturing partners throughout Asia. That creates a feedback loop where supply chain concerns compound investor anxiety about demand.

The question now is whether this is a temporary correction or the beginning of a more fundamental repricing. AI infrastructure spending has been treated as almost sacred – a bet you had to make regardless of near-term returns. But if investors start demanding clearer paths to profitability, the entire sector could face sustained pressure.

For Nvidia, the stakes are existential. The company’s valuation is predicated on continued explosive growth in AI chip sales. Any hint that growth is slowing – or that competition from AMD or custom chips from Google and Amazon is eating into market share – triggers outsized reactions.

The megacap bounce happening elsewhere in tech suggests this isn’t about broader market conditions. It’s a targeted reassessment of who wins in AI. And right now, investors are voting against the infrastructure layer.

The split between infrastructure and application-layer tech stocks marks a potential turning point in the AI investment narrative. After a year of throwing money at anything AI-adjacent, investors are finally asking who’s actually going to profit from this transformation. Chip makers and infrastructure providers built the foundation, but if the returns don’t materialize soon, they could find themselves funding someone else’s gold rush. The next few earnings cycles will be critical – particularly guidance around AI capital expenditure from the major cloud providers. If spending starts to moderate, the infrastructure trade could face a prolonged winter regardless of how revolutionary the technology actually is.