SemiAnalysis reports GCP revenue growth >100% YoY while Gemini struggles, arguing DeepMind's model failures fuel GCP's AI infrastructure boom.
Google Cloud Platform revenue is growing over 100% year-over-year, according to @SemiAnalysis_. The firm argues DeepMind's Gemini struggles long-term, yet GCP's AI infrastructure demand turns that into short-term gain.
Key facts
- GCP revenue growth >100% YoY per @SemiAnalysis_
- Gemini is Cooked but GCP is Cooking headline
- DeepMind's failure is GCP's short-term gain
- Google's Q1 2025 GCP revenue was $11.4B, up 28%
- Google committed $75B capex to AI infrastructure
Google Cloud Platform (GCP) revenue growth exceeded 100% year-over-year, according to @SemiAnalysis_. The firm's post — titled "Gemini is Cooked but GCP is Cooking" — argues DeepMind's long-term failure in the Gemini model line is paradoxically Google Cloud's short-term gain.
The thesis: even if Gemini lags rivals in model quality, the demand for AI compute and cloud infrastructure is so strong that GCP's AI-driven workloads are accelerating. The post does not disclose absolute revenue figures, only the >100% YoY growth rate. Google's last public earnings report showed GCP revenue at $11.4B in Q1 2025, up 28% YoY, so a >100% growth rate would represent a dramatic acceleration — the company has not confirmed this figure.
SemiAnalysis frames the dynamic as a structural hedge: Google's AI model team may be losing the benchmark race, but its cloud business is capturing the infrastructure spend that AI adoption requires. The firm's analysis suggests the market is rewarding GCP's position in AI data centers and TPU/GPU capacity over Gemini's model quality.
The strategic read
This is a contrarian take on the standard narrative that model quality determines AI winners. SemiAnalysis implies the real value is in the compute layer, where GCP's scale and TPU ecosystem matter more than Gemini's benchmark scores. If true, Google's $75B 2025 capex commitment to AI infrastructure [as previously reported] positions GCP to monetize regardless of DeepMind's model performance.
The long-term risk remains: if Gemini continues to lag, customers may shift workloads to rivals' models, but the infrastructure lock-in could persist. The post does not specify which benchmarks or metrics support the "cooked" claim about Gemini, nor does it provide GCP's absolute revenue numbers.
What to watch
Watch Google's next quarterly earnings call for GCP revenue disclosure — if the >100% YoY growth is real, expect the figure to appear in Alphabet's Q3 2026 results. Also track whether Google revises its 2026 capex guidance upward, which would signal continued infrastructure demand.
Originally published on gentic.news
Top comments (0)