I’m watching Alphabet (NASDAQ: GOOGL | GOOGL Price Prediction) ahead of first-quarter results due Wednesday, April 29, after the bell. Beyond cloud and ad headlines, this print is really about whether Google’s custom silicon story holds up against NVIDIA.
Custom Silicon Steps Into the Spotlight
The setup is loud. Shares are up 28.10% over the past month and nearly 119% over the past year, with the stock at $351.42 heading into the report. Last quarter delivered $113.83 billion in revenue, up 17.99% year over year, and EPS of $2.82, beating consensus by 7.22%.
What changed since February is the framing. Sundar Pichai positioned TPUs as a decade-long moat, citing the company’s seventh-generation Ironwood TPU alongside a 78% reduction in Gemini serving unit costs across 2025. Cloud accelerated to 48% growth, with backlog jumping 55% sequentially to $240 billion. Then came the eye-popping number: 2026 CapEx guidance of $175 billion to $185 billion.
What the Numbers Looked Like Last Year
| Metric | Q1 2025 Actual | Q4 2025 Actual | FY2025 Actual |
|---|---|---|---|
| Revenue | $90.23B | $113.83B | $402.84B |
| YoY Growth | +12% | +18.0% | +15.1% |
| Diluted EPS | $2.81 | $2.82 | $10.81 |
| Cloud Revenue | $12.26B (+28%) | $17.66B (+48%) | Run rate >$70B |
Note: Polymarket pricing implies a 95.4% probability of an EPS beat.
TPUs Are the Real Story This Print
I’ll be watching the TPU narrative more than the headline beat. Pichai already told us Google has “been investing in TPUs for over a decade” and that “just over half of our ML compute” is expected to go toward the cloud business in 2026. That is the line that reframes the Nvidia comparison. Google isn’t dependent on someone else’s roadmap. They are running their own.
You should focus on three things. First, capex absorption. With depreciation already up nearly $6 billion in 2025 to $21.1 billion, margin commentary matters more than the print. Second, cloud durability. The number of $1B+ deals in 2025 surpassed the previous three years combined, and Pichai expects supply constraints to persist all year. Third, Gemini monetization. Gemini Enterprise crossed 8 million paid seats four months after launch, and the Gemini App now serves over 750 million monthly active users.
The market is leaning bullish. Polymarket assigns 86% odds that Gemini MAU comes in above 800 million, and analysts hold 63 Buy ratings against zero Sells. The risk isn’t a miss. It’s whether management can defend the capex narrative without wobbling.
The TPU Argument Either Lands or It Doesn’t
This is the quarter where Google can prove the AI hardware race isn’t a one-horse contest. If Pichai shows TPU economics flowing into cloud margins and customer wins, the $175B to $185B capex commitment stops looking reckless and starts looking strategic. I think the setup favors a confident management tone. The numbers are the easy part. The TPU story is the one that decides the next leg.