Apple has already begun to prepare the introduction of generative AI in its products. During the presentation of the new iPad in London last Tuesday, the company led by Tim Cook recalled that since 2017 it has been introducing chips with a neural engine, a part of the system that processes data on its mobile phones, tablets and computers that is dedicated to artificial intelligence. Until now, this capacity served, for example, to greatly improve photographs, but putting a large AI language model inside a device is another matter.
On Tuesday, Apple announced that the new iPad Pros will arrive next week with a new chip of its own design, the M4. It was only six months ago that it presented the M3, with enormous capacity. Although it is true that the speed of renewal has been unusual, although there is a good chance that the M3 has not yet exhausted its life cycle. The main differences between the two are two: an image processor that handles the enormous amount of information necessary to operate its new screen with a double OLED panel and, of course, the incorporation of a large 10-core neural engine that gives it powerful in that field that the first PCs with AI chips would already like: 38 billion operations per second.
If you ask Apple spokespersons if, thanks to the new M4, the new iPad will be able to handle generative AI in its gut, without the need for a connection to the outside, the answer is, of course. Apple smiles because it trusts that it has already resolved the answer to the revolution in the technological world that at the end of 2022 brought about the arrival of ChatGPT. We will most likely know the next step within a month, when it announces its AI news during the WWDC developer conference.
In that call, Apple will explain how it will renew its different operating systems, which will reach users in the fall. By the time it is produced, it will already have a new generation of mobile phones, the iPhone 16, which will also be able to equip a chip with greater AI capabilities. The transition will also have two-speed tracks, because the older models will not have that special “brain.” The renewal of models among customers will also mark access to maximum capabilities, a gap situation that will not only affect Apple customers, but those of the entire industry.
In addition to this strategic move, according to The Wall Street Journal, Apple plans another fundamental move with the development of chips for servers, which are used in large data centers. According to this information, the Apple server chip would be intended for the execution of artificial intelligence models – a process called inference – instead of training AI models, where Nvidia has taken the lead and is the market leader with a great growth. In any case, it is a market in full explosion in which all the big players are looking to create a space for themselves.