In the world of technology, multibillion-dollar investments are the norm for large companies. But even in this context, Apple’s planned expenditure in the field of artificial intelligence for 2024 has managed to surprise us. According to the renowned analyst Ming-Chi Kuo, Apple is gearing up to invest a minimum of $620 million in AI servers in 2023, and the staggering amount of $4.75 billion in 2024.
The figures undeniably suggest a drastic shift in the company’s focus toward the development and implementation of AI. Kuo’s estimates indicate that Apple could acquire between 2,000 and 3,000 servers this year and up to 20,000 next year. These servers would be equipped with Nvidia’s HGX H100 8-GPU platform, described by the company as an artificial intelligence supercomputer. Each of these units is priced at around $250,000.
Now, why does Apple decide to make this investment? Kuo speculates that the company will choose to install and use its own servers to train large language models, instead of relying on virtual hosting from other cloud service providers. The reason: to enhance the security and privacy of data and developments.
According to Kuo, Apple might consider developing its own chips to reduce server costs in the future. As of now, however, the analyst claims he hasn’t seen any signs that Apple has decided to take this direction.
While Kuo expresses being “really worried” about the future of Apple’s generative AI if the company continues its current investment trajectory, the numbers presented indicate a positive change. One that we know involves Apple working on a smarter version of Siri and attempting to integrate AI into several of the company’s applications.
Indeed, the sum of nearly five billion dollars in AI servers sounds truly astronomical. However, it’s evident that such figures are necessary to compete in the rapidly advancing field of these technologies. We will see, as we await Apple GPT, how Apple’s investments in this field evolve and how AI develops within the company.