Google's AI chip business is getting harder for Nvidia to ignore (www.businessinsider.com)

Google's announcement of a partnership with Oracle to offer its Tensor Processing Units (TPUs) on Oracle's cloud marks a significant shift in the AI hardware landscape. This collaboration represents the first time Google's TPUs will be accessible on a major cloud service provider outside of its own infrastructure, enabling a broader adoption of its custom AI chips. The news also coincides with OpenAI's recent hiring for a software engineer to optimize its models for TPUs and its intention to purchase these chips from Google, further indicating a growing competition to Nvidia's dominance in the AI chip sector. This move is crucial for the AI/ML community as it opens the door for a multichip strategy among companies, which could dilute Nvidia's pricing power and market share. Analysts predict that Google's TPU revenue could surge to $100 billion by 2028, as companies increasingly diversify their AI hardware usage, reflecting a significant opportunity for Google after years of investment in AI silicon. This shift may not only bolster Google's position in the cloud computing and AI markets but also serve as a strategic counterbalance to potential downturns in its traditional search business amidst increasing competition from AI-driven technologies.