Alphabet shares rallied as Google rolled out its Gemini 3 model and Ironwood AI chips, signaling a sharper strategy that spans consumer, enterprise and infrastructure. The company is integrating generative AI across Search, YouTube and Cloud while pushing custom TPUs that partners such as Anthropic are adopting, challenging Nvidia’s dominance. Investors cheered the momentum—buoyed by a Berkshire stake and a rising market cap—even as analysts cautioned that the lead over OpenAI, Anthropic and others remains slim. Google’s vast YouTube corpus and rapid model cadence have become strategic assets, yet past stumbles—hallucinations and flawed image generation—underscore execution risk. The AI race is capital intensive; Big Tech expects 2025 capex to swell, and Google says it must double serving capacity every six months to meet demand. Nvidia still controls most AI silicon, arguing its GPUs are more flexible than Google’s ASIC TPUs, keeping the infrastructure contest tight. For Google, the path forward is clear but costly: scale the full stack while proving its models and chips translate into durable revenue.
Related article:
— Alphabet Investor Relations: earnings, filings and AI updates
