Apple's decision to use Google's Tensor Processing Units (TPUs) instead of Nvidia's popular GPUs for training its AI models signals a strategic shift that could impact the AI hardware market significantly. Despite Nvidia's dominance in the AI chip market, commanding around 80% of the market share, Apple opted for Google's TPUs, particularly the TPUv4 and TPUv5, to develop its foundational language models, known as Apple Intelligence. This choice likely reflects Apple's need for scalable and efficient infrastructure, leveraging Google's expertise in AI and data centers.Nvidia's market position remains strong, but Apple's move might indicate a growing trend among tech giants to explore alternatives as they seek specialised solutions for their AI needs. This could potentially challenge Nvidia's market dominance if other companies follow suit. However, it's also possible that Apple, known for its unique approach, is simply choosing the best available tools to achieve its goals, rather than signalling a broader industry shift.Nvidia's response and the overall market reaction will be crucial in determining whether this decision marks a turning point or is merely a one-off move by Apple to optimise its AI training process​.