AI Capex Reprices Markets, Corporate Growth Paths

Broadcom reported $16.7 billion in AI semiconductor revenue in a single quarter and projected approximately $58 billion for the full fiscal year, underscoring the scale of capital moving into AI infrastructure.
Applied Materials’ quarterly growth accelerated from a 2% decline in fiscal Q1 2026 to 11% growth in Q2 and 25% growth in Q3. CEO Gary Dickerson described AI-driven demand as “unprecedented,” while the company guided to roughly $10.25 billion in fiscal Q4 revenue.
Oracle said it had deployed 850 megawatts of capacity and shipped more than 300,000 GPUs operating at 97.9% utilization; customers were reportedly renewing GPU capacity at a 20% premium to prior contracts.
Meta’s quarterly results included a 27.96% increase in revenue to $60.80 billion, but diluted earnings per share came in at $6.18 versus the $7.22 analyst estimate. Legal charges of $2.4 billion and severance costs of $1.18 billion contributed to a 55% increase in total costs.
Applied Materials and Lam Research are exposed to the same underlying equipment-spending trend: AI increases the number of layers, process steps and materials needed per wafer, allowing equipment spending per wafer to rise even when chip volumes are flat. Lam is particularly concentrated in etch and deposition tools used in memory manufacturing.
The AI infrastructure boom is hitting a valuation wall as investors weigh massive capital spending against uncertain payback periods and higher interest rates. Despite record revenue growth at chipmakers and cloud providers—Broadcom posted $16.7 billion in AI semiconductor revenue in a single quarter—stocks in the sector have tumbled on worries that frontier-AI development may slow, reducing future demand for expensive chip factories Seeking Alpha.
The cash-flow gap is widening too. Meta's quarterly revenue jumped 27.96% to $60.80 billion, but earnings per share missed analyst targets, and total costs surged 55% due to legal charges and severance Seeking Alpha. Oracle built a $664 billion backlog while operating 850 megawatts of capacity, yet posted negative free cash flow amid massive capital expenditures Seeking Alpha.
Applied Materials' growth is picking up speed. Quarterly revenue growth jumped from a 2% decline in fiscal Q1 2026 to 11% in Q2, then 25% in Q3 Seeking Alpha. CEO Gary Dickerson called AI-driven demand "unprecedented," and the company guided to roughly $10.25 billion in fiscal Q4 revenue Seeking Alpha. Each AI chip requires more layers and process steps, which pushes equipment spending higher even when chip volumes stay flat.
Broadcom reported $16.7 billion in AI semiconductor revenue in a single quarter and projected approximately $58 billion for the full fiscal year Seeking Alpha. That figure alone shows how much capital is flowing into AI infrastructure. The sheer scale underscores investor confidence in near-term demand, even as longer-term returns remain uncertain Seeking Alpha.
Oracle deployed 850 megawatts of capacity and shipped more than 300,000 GPUs running at 97.9% utilization Seeking Alpha. Customers are renewing GPU capacity at a 20% premium to prior contracts, signaling strong demand. Yet Oracle posted negative free cash flow despite a $664 billion backlog and a rapidly growing cloud business, trapped between customer demand and spending needs Seeking Alpha.
Meta's quarterly revenue grew 27.96% to $60.80 billion, but diluted earnings per share landed at $6.18, missing the $7.22 analyst estimate Seeking Alpha. Total costs jumped 55%, fueled by $2.4 billion in legal charges and $1.18 billion in severance Seeking Alpha. The result: free cash flow plunged as spending and costs surged, despite topline strength. Investors now question whether AI investments will generate enough returns to justify elevated valuations.
Publishers
21
Articles
34
Reach
55