On the morning of September 10th, Beijing time, the Apple Autumn Launch Event, known as the "Technology Spring Festival", was held as scheduled, focusing on showcasing new products such as the new iPhone, Apple Watch, and Airpods. The iPhone 16, which is equipped with the latest 3nm A18 series chip, was the highlight of the event. Lei Zhiyong, Deputy Director of Equity Investment at Morgan Stanley Fund, believes that in this round of AI terminal innovation cycle, in addition to the growth in terminal sales, it is expected that the use of terminals will bring a huge demand for inference computing power, which is expected to further drive the demand for computing chips. At the same time, once AI is popularized on terminals, various software and model incomes may also see a real widespread adoption and outbreak. In terms of investment opportunities, he is optimistic about the pull of AI on cloud computing power and terminal consumer product demand. Apple's smart debut, will AI phones usher in a super cycle? Compared to the Apple Intelligence announced at the WWDC in June this year, this event did not see many changes in software, mainly including two new standalone apps represented by Writing Tools and Image Playground, a smarter Siri, and cloud server Private Cloud Compute emphasizing user privacy protection. In terms of hardware, the iPhone 16 series is equipped with a powerful A18 bionic chip, capable of meeting the computing power requirements of Apple Intelligence, with the machine learning speed of A18 doubling compared to the previous generation chip. Regarding the most anticipated launch time, Apple (AAPL.US) clearly stated that the English version of Apple Intelligence will begin to be available in the US next month, and will gradually launch in Australia, Canada, New Zealand, South Africa, and the UK from December onwards. Support for other languages such as Chinese, French, Japanese, etc., will have to wait until 2025. With the support of Apple Intelligence, the iPhone 16 is expected to release the pent-up demand and accelerate the replacement cycle of iPhones. According to market estimates, the shipment volume of the iPhone 16 series in 2024 is expected to reach 88-89 million units, and by the end of 2024, the global generative artificial intelligence (GenAI) phone market is expected to grow by 344%, capturing an 18% market share. Data source: IDC Lei Zhiyong, Deputy Director of Equity Investment at Morgan Stanley Fund and Manager of the Morgan Stanley Digital Economy Mixed Fund, previously stated that AI phones are very similar to the 5G replacement wave, and a 1% replacement rate is expected to boost Apple's sales by 6% and bring a 1.8% increase in Apple's revenue. However, compared to the replacement wave, Lei Zhiyong believes that the significance of Apple Intelligence lies in the fact that GenAI is expected to penetrate into the daily lives of the general public for the first time. Apple AI is expected to bring hundreds of billions of incremental revenue to Apple itself. If there are 300-500 million users paying 10 yuan for AI per month in 2025, this corresponds to nearly 5 billion yuan per month, while 1.5 billion existing users correspond to 15 billion per month, exceeding 100 billion per year. What's next for the AI industry? So what impact will the replacement cycle brought by the Apple Intelligence have on the AI industry, which has been under pressure recently? Recently, US technology companies, especially those related to AI chips, have been under pressure. Starting from the high point on July 11th, up to last Friday, September 6th, the Philadelphia Semiconductor Index has fallen by more than 23% (source: Wind). The successful holding of the Apple launch event has injected some vitality into the technology sector, with the Philadelphia Semiconductor Index rebounding by 2.15% on the day of the event, and the Nasdaq index also closing up 1.16% (source: Wind). In fact, compared to the somewhat "blurry" concepts of the metaverse and other emerging technologies, the tangible performance brought about by the rapid development of AI in the past two years is visible to the naked eye. Taking a leading global AI chip company as an example, its data center revenue related to AI has been steadily increasing over the past two years, with the data center revenue share of total revenue rising from less than 60% to 87.46% in the most recent fiscal quarter, all thanks to the real growth brought about by AI. Data source: Public report of a leading global AI chip company Therefore, despite the short-term pressure, the overall trend of artificial intelligence has not changed. Lei Zhiyong believes that the evolution of artificial intelligence is staged, and it may fall below expectations at a certain stage, but this does not prove that artificial intelligence has failed. The future of artificial intelligence is vast, aiming for general artificial intelligence, not just extracting information from the internet in a question-and-answer format. It should be able to replace humans in some difficult environments or repetitive tasks and help complete these tasks, but there is still a long way to go to achieve this goal. Lei Zhiyong also pointed out that while investment in artificial intelligence is already hot, from the industry perspective, artificial intelligence is still in its infancy stage. Specifically in terms of investment opportunities, Lei Zhiyong sees the pull of AI on cloud computing power and terminal consumer product demand as promising. He states that currently, giant companies at home and abroad are not only focusing on cloud computing power but also starting to evolve towards the application end, especially intelligent terminals. In the intelligent terminal industry chain, domestic companies can serve as supporting manufacturing chains, with orders serving as performance realization support. In this round of AI terminal innovation cycle, apart from the expected increase in terminal sales, he predicts that the use of terminals will bring a huge demand for inference computing power, which is expected to further drive the demand for computing chips. Moreover, once AI is popularized on terminals, various software and model incomes may also see a true widespread adoption and explosion.