SemiAnalysis Report: Gemini 3 Pro May Be Google's Model Peak; AI Focus Shifts to Cloud Infrastructure

icon MarsBit
Share
AI summary iconSummary
A new weekly market report from AI research firm SemiAnalysis suggests Gemini 3 Pro may be Google’s leading model. Leadership changes at DeepMind and Alphabet, including Demis Hassabis stepping back and Jeff Dean co-founding Discovery Loop, signal internal restructuring. The report estimates Gemini will trail behind OpenAI and Anthropic by 2026. Google Cloud is expected to benefit from cloud infrastructure and TPU commercialization. By 2027, third-party AI IaaS/TaaS ARR could exceed $730 billion, with TPU sales reaching $1.2 trillion.

Huo Xing Finance reports that on August 7, following personnel shifts at Google DeepMind, market focus on Alphabet’s AI narrative has shifted from model rankings to cloud business monetization. According to the latest report by AI market research firm SemiAnalysis, Demis Hassabis stepping back from day-to-day management of DeepMind, combined with Jeff Dean founding Discovery Loop alongside Sanjay Ghemawat, Quoc Le, and Oriol Vinyals, alongside Koray Kavukcuoglu taking over Gemini/DeepMind, signals a profound restructuring of Google’s frontier model team. SemiAnalysis assesses that Gemini 3 Pro may represent the peak of Google’s model competitiveness, with Gemini’s performance in 2026 clearly lagging behind competitors such as OpenAI and Anthropic. However, SemiAnalysis believes the biggest beneficiary of this restructuring will be Google Cloud. Previously, Gemini and GCP competed for compute resources; now, resource allocation is increasingly prioritized toward cloud business and TPU commercialization. SemiAnalysis estimates that Gemini’s annual recurring revenue (ARR) in Q2 2026 will be approximately $12 billion, while by the end of 2027, third-party AI IaaS/TaaS ARR on GCP could exceed $73 billion, with TPU sales reaching $120 billion. This also explains the recent market repricing of capital expenditures by large tech companies in AI: investors are concerned about the rapid cash burn of model-focused firms but remain willing to assign higher valuation certainty to companies capable of selling compute, cloud services, and accelerators. SemiAnalysis notes that GCP grew 82% last quarter, partly due to sales of TPU systems to external SPVs supporting data centers for clients such as Anthropic. As TPU backlogs expand, GCP’s 2027 growth rate may significantly exceed sell-side consensus and contribute incrementally to Alphabet’s earnings per share. SemiAnalysis believes the core narrative of Google’s AI story may be rewriting itself: if Gemini continues to fall behind, it will weaken Alphabet’s positioning in the frontier model race; however, if GCP can consistently convert TPU and compute demand into revenue, the market’s valuation anchor for Alphabet will increasingly shift toward cloud infrastructure, enterprise AI platforms, and compute supply capabilities.

Disclaimer: The information on this page may have been obtained from third parties and does not necessarily reflect the views or opinions of KuCoin. This content is provided for general informational purposes only, without any representation or warranty of any kind, nor shall it be construed as financial or investment advice. KuCoin shall not be liable for any errors or omissions, or for any outcomes resulting from the use of this information. Investments in digital assets can be risky. Please carefully evaluate the risks of a product and your risk tolerance based on your own financial circumstances. For more information, please refer to our Terms of Use and Risk Disclosure.