Google unveils TPU v7 AI chip & Q2 earnings: Cloud surge boosts after-hours stock 5%

On July 30 after the bell, Alphabet, Google's parent, released Q2 2026 earnings, also announcing the new AI chip TPU v7. Driven by earnings beat and chip breakthrough, Google's after-hours stock rose over 5% to $212.30. Analysts noted AI business expansion is becoming Google's new growth engine.

Earnings Highlights: Cloud leads, ad revenue stabilizes

Earnings data showed Alphabet Q2 total revenue reached $98.2B, up 14% YoY, beating estimates of $95.7B; net profit $21.8B, up 18% YoY, EPS $1.72, also above Wall Street expectations. Google Cloud was the standout, revenue up 32% YoY to $12.8B, accelerating for three consecutive quarters. Cloud profitability also improved, with operating profit jumping from $0.31B to $1.25B, marking the segment's profit ramp-up. Advertising, despite macro pressures, grew 8% YoY to $63.7B, driven by search and YouTube. Other bets like Waymo and Verily remain in investment phase, losing about $1.5B combined.

TPU v7: Performance leap targeting AI training and inference

The TPU v7 is Google's latest custom AI chip. It uses 3nm process with over 200 billion transistors, delivering 3x compute performance over TPU v6. For large AI model training, TPU v7 clusters can cut training time by over 60%; for inference, energy efficiency improves 40%. CEO Sundar Pichai said on the call: "TPU v7 will redefine AI infrastructure cost and efficiency; we have initial cooperation intentions with several top AI companies." TPU v7 is expected to be available to enterprise customers via Google Cloud in Q4 2026. Google also revealed development of TPU Edge series for edge computing, expanding its AI chip footprint.

Industry view: AI arms race escalates

As Google launches TPU v7, the global AI chip race intensifies. Nvidia dominates AI training with GPUs, while AMD, Intel and many startups enter. Google's TPU iteration accelerates: v5 to v6 took 2 years, v6 to v7 only 1.5 years, showing determination to catch up and surpass. Analysts note TPU v7 can match Nvidia's latest Blackwell Ultra on certain AI workloads, and Google's self-use + cloud model gives cost advantages. However, Nvidia's CUDA ecosystem retains developer stickiness, so Google will struggle to unseat it soon. For investors, cloud growth and AI chip breakthrough boost valuation, but rising capex may erode profits.

Technical analysis: Key resistance and support levels

Technically, Google stock pulled back from its record high of $216 in early June, closing near $202 on July 29. After-hours surge above $210 recaptured the 20-day moving average. Short-term resistance at the previous high of $216; if broken, new upside may open. Support at $200 round number and 50-day MA near $195. MACD histogram green bars expanded, DIF crossed above DEA, showing bullish alignment. Volume spiked after hours, indicating positive reaction. Investors should watch cloud growth and TPU order conversion; consider positioning on pullbacks to support.

Market outlook: AI monetization remains core narrative

Overall, Google's Q2 earnings validate AI monetization. Cloud's high growth shows enterprises accelerating cloud migration, with AI services as a key driver. TPU v7 strengthens the full-stack AI narrative from chip to model to applications. However, risks remain: macro uncertainty may impact ad budgets; high AI chip R&D costs compress margins short-term; antitrust pressure - DOJ's search monopoly lawsuit continues. For growth investors, focus on cloud revenue share trend and buy on dips; for value investors, wait for more attractive valuations. Google's forward P/E of ~27x is near its 5-year average; with AI dividends, long-term value is evident.