Sundar Pichai and Tim Cook
Source: Reuters; apple
apple said on Monday that the artificial intelligence models that power Apple Intelligence, its AI system, were pre-trained on processors designed by Google, a sign that the tech giants are looking for alternatives Nvidia when it comes to cutting edge AI training.
Apple’s choice for of Google The domestic Tensor Processing Unit (TPU) for training was described in detail a technical paper just published by the company. Separately, Apple released a preview version of Apple Intelligence for some devices on Monday.
Nvidia’s expensive graphics processing units (GPUs) dominate the market for high-end artificial intelligence training chips and have been in such high demand over the past two years that it has been difficult to supply the required quantities. OpenAI, Microsoftand Anthropic all use Nvidia GPUs for their models, while other tech companies, including Google, After, Oracle and Tesla they buy them to build their AI systems and offerings.
Meta CEO Mark Zuckerberg and Alphabet CEO Sundar Pichai made comments last week saying their companies and others in the industry may be overinvesting in AI infrastructure, but acknowledged the business risk is too high .
“The downside of being behind is that you’re out of position for the most important technology for the next 10 to 15 years,” Zuckerberg said in a podcast with Bloomberg’s Emily Chang.
Apple doesn’t name Google or Nvidia in its 47-page paper, but noted that its Apple Foundation Model (AFM) and AFM server are trained on “Cloud TPU clusters.” This means that Apple rented servers from a cloud provider to perform the calculations.
“This system allows us to train AFM models efficiently and at scale, including AFM-on-device, AFM-server, and larger models,” Apple says in the paper.
Representatives for Apple and Google did not respond to requests for comment.
Apple was set to reveal its AI plans later than many of its peers, which strongly embraced genetic AI soon after OpenAI’s ChatGPT was released in late 2022. On Monday, Apple unveiled Apple Intelligence. The system includes several new features, including a refreshed look for Siri, better natural language processing, and AI-generated summaries in text fields.
Over the next year, Apple plans to roll out features based on genetic artificial intelligence, including image creation, emoji creation, and an enhanced Siri that can access a user’s personal information and take actions within apps.
In Monday’s paper, Apple said the AFM in the device was trained on a single “piece” of 2048 TPU v5p chips that work together. This is the most advanced TPU, first released in December. The AFM server was trained on 8192 TPU v4 chips configured to work together as eight pieces over a data center network, according to the paper.
Google’s latest TPUs cost less than $2 per hour the chip is used when reserved three years in advance, according to on the Google website. Google first introduced its TPUs in 2015 for internal workloads and made them publicly available in 2017. They are now among the most mature custom chips designed for artificial intelligence.
However, Google is one of Nvidia’s top customers. It uses Nvidia GPUs and its own TPUs to train AI systems, and also sells access to Nvidia’s technology in its cloud.
Apple previously said that inference, which means taking a pre-trained AI model and running it to generate content or make predictions, would happen in part on Apple’s own chips in its data centers.
This is the second technical paper on Apple’s AI system, following the publication of a more general version in June. Apple said at that time that she used TPU as she developed her AI models.
Apple is scheduled to report quarterly results after the close of business on Thursday.
