When ChatGPT launched in 2022, Google was caught off guard, but the release of Gemini 3 and the Ironwood AI chip this month signals a major comeback for Alphabet. Ironwood, Google’s seventh generation tensor processing unit (TPU), allows customers to run large, data-intensive models efficiently, while Gemini 3 delivers smarter answers with less prompting than previous AI models.
Salesforce CEO Marc Benioff praised Gemini 3, switching from daily ChatGPT use after experiencing its performance. Alphabet’s stock surged over 5% on Monday, adding to last week’s 8% gain, driven by investor confidence, including Berkshire Hathaway’s $4.3 billion stake. Alphabet shares are up nearly 70% this year, outperforming competitors like Meta.
Google CEO Sundar Pichai appears to have aligned its AI offerings, serving both consumers and enterprises effectively. After initial missteps with products like Imagen 2 and AI Overviews, Google refined its models and launched Gemini 3 rapidly following Gemini 2.5. The hyper realistic image generation tool Nano Banana, topping the Apple App Store, further highlights Google’s AI strength. YouTube’s vast content library gives Google an edge in training AI for image and video tasks.
Google has also integrated AI across enterprise products, driving cloud revenue, which reached $100 billion in the latest quarter. Ironwood TPUs are nearly 30 times more power-efficient than Google’s first TPU, giving it leverage in AI chip deals and enterprise partnerships.
Despite these wins, Google faces stiff competition from OpenAI, Meta, and others, as frontier AI models remain closely matched. The AI race is costly, with companies collectively spending over $380 billion this year on development and infrastructure. Google must double its serving capacity every six months to meet AI demand, while Nvidia still controls over 90% of the AI chip market.
While Google’s AI comeback is impressive, maintaining a lead will require ongoing investment and innovation in both consumer and enterprise AI offerings.
