Apple made an announcement on Monday regarding its AI system, Apple Intelligence, revealing that the artificial intelligence models behind it were pre-trained on Google’s Tensor Processing Units (TPUs) instead of Nvidia processors. This decision by Apple sheds light on the growing trend among major tech companies to explore alternatives to Nvidia when it comes to cutting-edge AI training.
The revelation that Apple chose Google’s TPUs for training its AI models indicates a shift in the industry dynamics. While Nvidia’s GPUs have traditionally dominated the high-end AI training chip market, companies like Apple, Meta, Oracle, and Tesla are now looking beyond Nvidia for their AI infrastructure needs. This move suggests that there is a growing demand for diverse and cost-effective options in the AI training chip space.
In a recently published technical paper, Apple detailed its use of Google’s TPUs for training the Apple Foundation Model (AFM) and AFM server. By leveraging Cloud TPU clusters for training, Apple aimed to achieve efficient and scalable model training, including on-device and server-based AFM models. This strategic partnership between Apple and Google underscores the collaborative efforts within the tech industry to advance AI infrastructure.
Apple’s late entry into the generative AI landscape, following the launch of ChatGPT by OpenAI in late 2022, raises questions about the company’s competitive position in the AI domain. However, the introduction of Apple Intelligence with enhanced features such as Siri improvements, natural language processing enhancements, and AI-generated summaries signals Apple’s commitment to catching up in the AI race. Moreover, Apple’s roadmap for incorporating generative AI functions like image and emoji generation into its offerings demonstrates a strategic shift towards AI-driven user experiences.
Google’s development and deployment of TPUs for AI workloads signify the company’s leadership in custom chip design for artificial intelligence. The availability of Google’s TPUs at a competitive price point underscores the accessibility and scalability of AI training resources for tech companies. Google’s dual reliance on Nvidia’s GPUs and its own TPUs for training AI systems exemplifies the collaborative ecosystem among tech giants in advancing AI capabilities.
Apple’s decision to utilize Google’s TPUs for training its AI models reflects a broader trend in the tech industry towards diversifying AI infrastructure providers. By embracing Google’s TPUs, Apple aims to enhance the performance and scalability of its AI system while leveraging cost-effective resources. As the competition in the AI landscape intensifies, strategic partnerships and technology collaborations will play a crucial role in shaping the future of artificial intelligence.