In recent years, Apple has actively promoted its chip self-development layout strategy, starting from the A series and M series processors. The Apple 16e launched in the first half of this year was the first to introduce Apple's first 5G network data chip C1, as well as theiPhone AirThe N1 wireless network chip and the upgraded 5G networking data chip C1X not only allow Apple to better control hardware performance and energy efficiency, but also lay the foundation for future "Apple Intelligence" services and device-side AI layout.
OnEarlier interviewAt the time, Apple's vice president of platform architecture, Tim Millet, and vice president of wireless technologies, Arun Mathias, pointed out that with the introduction of the new N1 wireless network chip and the upgraded C1X 5G network data chip in the iPhone Air, Apple has greater control over the device's operating behavior.
Tim Millet said: "When we have control, we can do things that we can't do by buying ready-made chips." This is different from the 2019Acquired Intel's wireless network businessAfter that, the strategy of gradually getting rid of external supplier Qulcomm was consistent.
However, AppleiPhone 17 and iPhone 17 Pro series modelsIt still maintains its cooperation with Qualcomm and adopts the same Snapdragon X80 5G network data chip as the Snapdragon 8 Elite. This shows that Apple has not introduced its own network data chip design into mainstream models as the market speculated. It may be to avoid the self-made network data chip affecting the user experience of the main sales models.
Arun Mathias explained that the new C1X 5G network data chip is not only faster thaniPhone 16eThe C1 used is twice as fast and consumes 30% less power than the previous generation Qualcomm 5G data chip. By designing its own wireless network chip, Apple can more efficiently integrate Wi-Fi signal positioning capabilities, allowing devices to locate without frequently waking up the main processor, thereby reducing power consumption and extending battery life.
Even though the iPhone 17 and iPhone 17 Pro series still use Qualcomm to provide 5G networking data chips, the inclusion of the C1X chip and N1 chip in the iPhone Air further highlights Apple's use of more core self-made components to test the waters, thereby gradually taking control of key components into its own hands.
A19 Pro brings GPU neural accelerator, a significant leap in AI computing power
In addition to improved communication performance, the updated A19 Pro processor is the core highlight of this new iPhone evolution. Tim Millet stated that Apple has not only enhanced the Neural Engine but also integrated the Neural Network Accelerator directly into the GPU core, further accelerating machine learning and AI inference through the GPU, even reaching MacBook Pro-level machine learning computing performance.
Tim Millet emphasized that Apple is building "the industry's best on-device AI capabilities," focusing not only on privacy and security but also on ensuring real-time and energy-efficient computing. The new GPU architecture can switch between traditional 3D rendering instructions and neural network computing, creating a new category of "programmable computer" for the iPhone. This means that future iPhones will be able to simultaneously handle both game graphics and AI workloads, and even allow more generative AI features to run directly on the device.
Although "Apple Intelligence" has not yet demonstrated a complete conversational AI experience like ChatGPT, with the comprehensive upgrade of its own chips in communication, graphics, and neural computing, Apple has clearly completed the underlying hardware layout and is ready for the next stage of AI technology competition.
