Apple is seriously tackling the field of artificial intelligence (AI) by collaborating with Broadcom to create its first chip intended for servers in its data centers. This initiative marks a crucial step for the Cupertino company, which aspires to reduce its dependence on Nvidia while strengthening its technological autonomy.

A strategic partnership with Broadcom

According to The Informationthis partnership takes place within the framework of the internal project known by the name of code Project ACDC (Apple Chips in Data Center). In May, The Wall Street Journal had already highlighted the need for Apple to catch up in AI. This project is part of a broader strategy aimed at enriching the brand’s product portfolio by integrating innovative AI functionalities.

The integration of ChatGPT into the Apple ecosystem

This week, Apple announced the integration of ChatGPT in its operating systems iOS 18.2, iPadOS 18.2 And macOS Sequoia 15.2. Thanks to this combination, the voice assistant Siri is enriched with the response capabilities of the OpenAI chatbot, thus improving the user experience.

At the same time, the brand offers other tools focused on intelligence, such as:

• Image Playground : a visual creation tool
• Genmoji : a dynamic writing system
• Visual Intelligence : available on iPhone 16 models

Although these tools may seem gimmicky, they reveal significant energy consumption, prompting Apple to optimize its resources.

Towards high-tech manufacturing

The new AI chip, named Baltrais expected to be ready for mass production by 2026. In this context, Broadcom will supply chiplets and make its infrastructure expertise available. Indeed, AI servers require highly connected processors, an area where Broadcom excels.

TSMC’s N3P engraving: A technological breakthrough

For its manufacture, Apple plans to use the advanced engraving process N3P of TSMC. This process focuses on optimizing transistor density, enabling component miniaturization while maximizing performance and simultaneously reducing power consumption.

This advance responds to certain limitations of the previous process, N3Ewhich although effective in reducing production costs, had a negative impact on transistor density.

Reduce dependence on Nvidia

With this initiative, Apple joins other tech giants, such asOpenAIwhich are also developing their own chips to power AI services. This approach aims to reduce dependence on expensive and rare processors such as those of Nvidia.

At its annual developers conference in June, Apple discussed its desire to use its own server chips to support AI features on its devices. This orientation is part of a guideline that the company has already taken successfully, particularly with its processors. Mwhich replaced the chips Intel.

In conclusion, the development of this AI chip by Apple represents a significant advance in the technological landscape, marking a new era of innovation for the firm.