Apple's baseband dream may really run into a big problem.
It's no secret to consumers that Apple has struggled to maintain its independence for years. In order to reduce dependence on companies such as Qualcomm or Broadcom, Apple has long formulated a plan for self-developed baseband chips, and previously laughed at Qualcomm and used Qualcomm baseband just because the self-developed baseband is not yet mature and requires transitional components.
The problem is that it seems that the development of Apple's self-developed baseband does not seem to be going very well. Not only that, Tianfeng International analyst Guo Mingxi said a few days ago that Apple has suspended the work of Wi-Fi chips under development, which also means that Apple's Wi-Fi supplier - Broadcom will continue to provide Wi-Fi chips for Apple in the foreseeable future, including chips for the iPhone 15/15 Pro series models to be released in 2023.
Source: twitter@ Guo Mingxi
The serious brain drain seems to have plunged Apple's chip division, which was once invincible, into the mud. Not to mention the A-series and M-series chips whose performance is standing still, even the baseband chips and Wi-Fi chips that were planned to go online early have not been progressed. The question is, why did Apple's self-developed baseband fall short? Why is Apple so concerned about self-developed baseband chips?
Apple's baseband dream
Looking back carefully, in the matter of baseband, Apple seems to have always had a hurdle that cannot be bypassed.
In fact, since the original iPhone, Apple's signals have been known for being unstable. Among them, as a smartphone that claims to "change everything", the iPhone 4, which first entered the country, was due to design errors, and frequently appeared that the signal was sharply attenuated or even lost when the user held it, which was also called "death grip" by netizens at the time.
Source: imore
The emergence of "Death Grip" has promoted Apple's supply chain upgrade, starting with the iPhone 4s, Apple replaced the baseband supplier with Qualcomm. Not to mention, Qualcomm's baseband is indeed level, which makes the generations of products from iPhone 4s to iPhone 6s basically have no signal-related problems, and Apple naturally maintains the position of Qualcomm baseband as the number one customer.
It's just that although the Qualcomm baseband is good, the price is not cheap at all. In addition to the cost of the chip, Qualcomm will also take a certain percentage of the patent share according to the price of the mobile phone, which has also caused strong dissatisfaction among domestic and foreign mobile phone manufacturers, and various manufacturers have successively started patent wars with Qualcomm. Affected by this, since the iPhone 7 series, Apple began to mix Qualcomm and Intel baseband, and then completely turned to the embrace of Intel baseband.
Source: Apple
Embarrassingly, the performance of Intel's baseband is really a bit crotch-pulling, and the power consumption, frequency band, signal strength, and uplink and downlink speed are all hit by Qualcomm at the same time, which also makes iPhone XS, iPhone XS Max and iPhone XR three products fall into the "signal gate", Apple's product reputation almost plummeted.
If the 4G era can be tolerated, after entering the 5G era, because of the process crotch, the XMM 8160 5G baseband released by Intel can only be built using 10nm process technology, which is not only inferior to Huawei's Baron 5000 baseband (7nm process) in terms of performance, but also delays in mass production due to various problems, which also leads to Apple finally having to spend money and Qualcomm to settle to ensure the normal listing of the iPhone 12 series.
If you think Apple will soften on Qualcomm, you're dead wrong. Shortly after the settlement with Qualcomm, Apple announced the acquisition of Intel's baseband business for $1 billion, once again demonstrating its determination: not to let Qualcomm make so much money from the iPhone. Since Intel can't do it, let Apple do it itself to study the solution.
Source: 95Mac
After Apple announced its self-developed baseband, it can be described as a succession of successive successive authoritative media that Apple's self-developed baseband will appear in 2023, even Qualcomm, which provides Apple's baseband, said at its own investor day event that it is expected to supply only 20% of Apple's baseband chips in 2023, and the currently used Broadcom Wi-Fi and Bluetooth chips will be fully replaced.
However, Guo Mingxi's recent tweet completely broke the previous rumors. Apple not only has the baseband chip stagnated, but even the research and development progress of Wi-Fi chips is not as expected, and the combination of Qualcomm + Broadcom will continue to provide Apple with 5G baseband/Wi-Fi chips in the foreseeable future. Getting rid of the dependence on Qualcomm + Broadcom components is not a simple matter for Apple.
Baseband requires long-term accumulation
The question is, why does Apple's chip department, which has repeatedly achieved great results in core SoCs, keep hitting a wall on the baseband?
In my opinion, the reasons are mainly due to two points.
First of all, the centralization of communication patent technology makes it difficult for these manufacturers to achieve breakthroughs. If you want to develop a communication baseband with independent property rights, you must obtain relevant communication patent authorization, and the patents of these communication protocols are in the hands of communication operators such as Qualcomm, Huawei, Ericsson, and Nokia. The baseband giant Broadcom in the era of functional machines withdrew from the baseband business in the 4G era because of the lack of patented 3G and 4G communication technologies.
Source: Apple
Secondly, there is the lack of empirical technology. In order to develop communication basebands with independent property rights, manufacturers must not only have relevant knowledge of semiconductor chip design, but also master the underlying knowledge of mobile communication systems. For example, how mobile phones can connect and transmit with different base stations in the process of moving, how to increase power to improve signal capabilities in remote environments, and how to do compatibility and adaptation with hundreds of network operators, the adjustment of these products requires a large amount of commercial data.
From the current situation, the process of Apple's self-developed baseband is obviously not as smooth as that of self-developed SoCs, and it can even be said that it is not clear enough. However, from Apple and Qualcomm to court, to shake hands with Qualcomm, to the acquisition of Intel's baseband division, this farce has only been two or three years. Lack of time, lack of patents, lack of experience are the key issues that the team needs to solve at the moment.
The purpose is to open source throttling
In the end, there is another question left for us: why is Apple so concerned about self-developed baseband chips?
For Apple, open source throttling is definitely a key point. According to the BOM cost analysis of the iPhone 14, despite the sharp rise in the cost of A16 Bionic, the Snapdragon X65 5G baseband is still the most expensive component, with a procurement cost of nearly $100. Apple's self-developed baseband means that it can reduce comprehensive costs, bypass Qualcomm's patent barriers, and avoid being "stuck in the neck" and "charge protection fees" by Qualcomm.
Not only that, self-developed baseband means that Apple can develop at its own pace and better adapt to its own products and functions. Now the internal space of smartphones is expensive, and the external Qualcomm baseband is never a long-term solution. It is foreseeable that Apple will eventually move towards the road of integrated self-developed SoCs to avoid external basebands occupying more motherboard space, which is also helpful for heat control and power saving.
Source: Intel
Of course, for users, there is only one very simple thing they care about: "If you use your own baseband one day, will Apple be able to improve the iPhone's signal problems?" ”
The answer may not be the same as everyone imagines, first of all, Apple's self-developed baseband is inherited from the research results of Intel's baseband team, and Intel's baseband products can be said to be notoriously "outdated" in the market. Considering that Apple itself is a newcomer in the communication market, coupled with the "crane tail" in the market of Intel's baseband department, it is somewhat foolish to develop a self-developed baseband product that surpasses Qualcomm baseband in a short period of time.
Secondly, there are many factors that affect the mobile phone signal, in addition to the baseband, hardware factors such as RF system, antenna performance, design of the whole machine, antenna arrangement, as well as software optimization and system BUG may directly affect the signal. For example, even if Apple replaces the Qualcomm baseband, the iPhone 13 and iPhone 14 series still have signal problems.
In other words, Apple's self-developed baseband is not to improve the iPhone signal, but to improve the only way under the general trend of hardware profit margins. It's just that from the current progress, the difficulty of self-developed baseband is much more difficult than that of self-developed chips, which means that more time, energy and manpower will be spent, and the current setback is the only way to self-develop chips and control the core industrial link. As the company's strategic orientation, it is only a matter of time before Apple uses its own baseband.
It's just the end result... Xiao Lei can only wait and see.