At a Glance
Yoon Jae-ho, OneDayTrading Editorial Board
Alphabet's $44.9 billion in second-quarter capital expenditure signals that AI competition has shifted from software updates to a battle over data centers, servers, networking, and power. The fact that this year's AI investment has grown to roughly 303 trillion won reads as an order book for Nvidia and memory makers—and a depreciation bill for Alphabet shareholders.
Revenue of $119.8 billion topped the LSEG consensus estimate of $116.93 billion. However, adjusted EPS of $2.85 fell short of the $2.89 expected. What the market should focus on isn't the growth rate, but the speed at which AI capital spending converts into profit.
Why It Matters Now
AI investment isn't a one-time server purchase. It comes bundled with GPUs and TPUs, HBM, high-speed networking, cooling systems, and power contracts all at once. Alphabet's second-quarter capital expenditure rose 100% year-over-year and roughly 26% from the first quarter's $35.7 billion. At this pace, the cost structure of a search-advertising company is effectively shifting toward that of a cloud infrastructure company.
The first thing investors need to translate is the supply chain. Alphabet's disclosure of 950 million monthly active users on the Gemini app and 22 billion tokens processed per minute via API signals that inference demand has entered the live-service stage. This is no longer a phase requiring only large training clusters—it's a stage where inference servers are needed repeatedly as the user base grows. Nvidia, AMD, Broadcom, and HBM suppliers all have direct exposure to this demand.
For Alphabet itself, though, the math is different. Even as AI integration into cloud and search lifts revenue, depreciation and power costs hit the income statement first. That structure is directly related to why EPS came in slightly below consensus. The tech narrative remains strong, but ultimately the stock price wants to confirm how much gross profit each dollar of investment generates.
Frequently Asked Questions
- Were Alphabet's earnings good? Revenue of $119.8 billion beat expectations. However, EPS of $2.85 came in slightly below expectations, leaving a cost overhang.
- Why does expanding AI investment matter for semiconductors? Data center expansion translates into orders for GPUs, HBM, networking chips, and power semiconductors. As service usage grows, demand for inference equipment recurs as well.
- Is this an unqualified positive catalyst for Alphabet? No. It's positive if cloud growth and AI-driven ad monetization accelerate, but margins get squeezed if depreciation and power costs rise faster.
- What does guidance for higher capital expenditure next year imply? It means AI demand is being treated as a multi-year investment plan rather than a short-term trend. At the same time, the window for verifying cash flow returns also lengthens.
Related Stocks and Sector Impact
- Alphabet: Integrating AI into search, YouTube, and cloud strengthens revenue resilience. However, massive capital expenditure pressures near-term EPS and free cash flow.
- Nvidia: Sustained capex from hyperscalers supports demand for AI accelerators. The key point is the absence of any signal that investment is slowing.
- AMD: Demand for alternative data center accelerators remains open. The more Nvidia's dominance is seen easing, the greater the upside.
- SK Hynix (000660) / Samsung Electronics (005930): Demand for HBM and high-value-added DRAM is likely to continue. What matters more than price is yield and the pace of customer qualification.
- Broadcom: This is a period of rising demand for custom AI chips and networking ASICs. Google's expanding TPU ecosystem is favorable for custom silicon providers.
Investment Considerations
- In the next earnings report, watch both cloud revenue growth and operating profit margin together. If revenue rises while margins decline, the stock (ticker) will price in the cost burden first.
- Even if capex guidance is raised, semiconductor stocks don't always rally. If share prices have already priced in the order increase, returns will be muted.
- Paid conversion and ad pricing matter more than the sheer number of AI service users. It's worth checking whether Gemini traffic is cannibalizing search advertising or reinforcing it.
- Rising interest rates lower the present value of long-term capital investment. In a Big Tech capex-driven market, U.S. 10-year Treasury yields need to be watched alongside it.
Overall Outlook
Alphabet's numbers point less to a cooling AI cycle and more to a shift into a heavier-cost phase. The bullish scenario is clear: if Gemini usage and cloud demand keep growing, and the roughly 303 trillion won in investment is recouped through search, cloud, and developer-platform revenue, Alphabet will prove it has the financial strength to absorb AI infrastructure costs. In that case, profit estimates across the AI chip and HBM supply chain would also hold up.
The bearish trigger is equally clear. If capital expenditure keeps rising while cloud growth slows or EPS continues to miss consensus, the market will re-price AI not as a growth investment but as a margin-destroying cost. The next checkpoints are Alphabet's third-quarter cloud revenue, the pace of capex execution, and order commentary from Nvidia and memory makers' customers. The competition in AI has shifted from how many models a company can ship to how much cash the hardware running those models can generate.
This article was automatically summarized and analyzed based on the original news report. View original (Yonhap News, Securities)
