Apple Favors Google Chip Power to Train Its AI Models 

Apple’s AI models for Apple Intelligence were pretrained using processors designed by Google, exploring alternatives to Nvidia’s for advanced AI training. 

According to a technical research paper published by Apple on Monday, the company opted for Tensor Processing Units (TPUs), known to be designed by Google, for the training of their AI models. This decision shows that the iPhone maker is trying to diversify its technology sources. 

Apple’s New Training Ground 

Apple’s 47-page technical document never mentioned Google or Nvidia. It did, however, point out that the Apple Foundation Model and AFM server were trained on “Cloud TPU clusters.” In other words, Apple rented cloud servers outfitted with Google’s TPUs to train their AI models. 

Apple explained, “This system allows us to train the AFM models efficiently and scalably, including AFM-on-device, AFM-server, and larger models.”  

In addition, the company further noted in the paper that AFM on-device was trained on a “slice” of 2048 TPU v5p chips. The AFM-server was trained with 8192 TPU v4 chips configured to work together as eight slices through a data center network. 

What Are TPUs? 

Tensor Processing Units (TPU) are chips designed by Google to accelerate the tasks of AIs. These developed chips are to handle complex mathematical calculations required for AI tasks, for example, image recognition or natural language processing. 

 Such computations are carried out by TPUs at a much faster and more efficient speed compared to regular computer chips, enabling quicker and more effective training and making predictions of AIs. 

On the other hand, Apple is set to launch several new features within its Apple Intelligence, such as a new interface for Siri, advanced natural language processing, and AI summaries of text. Such enhancements show Apple’s commitment to deeply integrating AI into its products. 

Final Thoughts 

Apple’s AI models decision to choose Google’s TPUs instead of Nvidia’s GPUs for its AI training represents a major shift in the technology landscape. With companies diversifying their technology sources, the competition in the AI industry is becoming more serious. 

Though, growth of the industry is important, and diversification is necessary. Relying on one provider might have implications for the process of innovation and might also create somehow restricted access.  

In adopting varied technologies, companies reduce risks and foster a more robust and dynamic AI ecosystem. This move by Apple also underscores an important point about flexibility and adaptability in tech strategy.

Inside Telecom provides you with an extensive list of content covering all aspects of the tech industry. Keep an eye on our Tech sections to stay informed and up-to-date with our daily articles.