For Apple, the signal issue seems to be an insurmountable hurdle that it can never get around.

As early as the iPhone 4 in 2010, there was a "Antenna Gate" incident. On June 24th of that year, the world's attention was on the official release of the iPhone 4. However, just a few hours later, an American netizen posted on a forum, claiming that the frame antenna design of the iPhone 4, which the company was proud of, had a fatal flaw. When users held the iPhone 4 tightly with their hands, the mobile network signal would decay to an incommunicable level within a few minutes, which is known as the "Death Grip."

However, Apple still maintained its arrogance and self-righteousness, which has been a constant throughout the history of Apple. A loyal user sent an email to Steve Jobs, saying, "I really like this phone, but when I hold the metal seam on both sides of the phone, the signal is gone. Can it be solved?"

Advertisement

But Jobs refused to admit the problem. He even told users, "You'd better change the way you hold the phone." Even at the "Antenna Gate" press conference, he avoided apologizing and simply said, "You know, we are not perfect, and neither are mobile phones. But we hope all users are satisfied. If you don't know this, then you don't understand Apple."

In the end, Apple did not recall these problematic iPhone 4s, but instead gave away a free Bumper phone case, and even shifted the blame to other mobile phone manufacturers such as Nokia, Motorola, and BlackBerry, claiming that their smartphones would also have such problems. It was later regarded as a model of crisis public relations.

But now, the signal of the iPhone is still the most complained about point by users. Everyone has more or less a doubt in their hearts: Apple, which has created the world's strongest mobile chip, why can't it solve the signal problem, and that inconspicuous baseband chip?

The indispensable baseband chip

The first-generation iPhone in 2007 was not as mature as it is today. Apple, which had never made a mobile phone, used its existing iPod and Mac supply chain to "assemble" a hardware device: the processor, NAND, and SDRAM were from Samsung, the touch screen came from the German Balda, the image sensor came from Micron, MARVELL was the WiFi chip, Wolfson was responsible for the audio chip, and even Intel participated, providing NOR and SRAM chips for Apple...

As for the indispensable baseband, Apple chose Infineon, the semiconductor department of the German giant Siemens, model M1817A11. According to French analysts at the time, Apple used Infineon's MP-E+ or MP-EU technology platform. Both platforms can manage Flash, Java, and MMS, and also have built-in GPS, FM tuner, and 3G functions. At that time, some optimistic netizens even predicted that shortly after the release, the first-generation iPhone (only supporting 2G network) could support 3G network through firmware updates.Infineon's communication solutions division had already seen a decline in net sales in 2006 due to the reduced demand for baseband products and pricing pressure, as well as the bankruptcy of its main customer, Ben-Q. The company began to cut costs by laying off employees. However, after winning the big customer, Apple, Infineon's gloomy prospects were instantly cleared up. After the release of the iPhone teardown report, Infineon's stock even rose by 3% on July 2 of that year, reaching $17.03.

Logically, Apple's help in resolving Infineon's urgent communication business should have been a great joy, but Infineon's tight brows did not ease at all, but instead fell into a dilemma.

The problem still lies in Infineon's baseband products. The first-generation iPhone did not support the widely popular 3G, which means it was already outdated just after its launch. It was an urgent need for consumers and Apple to quickly release a 3G model. However, in fact, as a European company, Infineon did not have much experience with the U.S. 3G network and was still in the stage of feeling its way.

The iPhone 3G was officially released in July 2008. Due to its support for the faster 3G network, it became one of the best-selling mobile phones on the market, with sales of 1 million units in the first week. However, its performance on the mobile network was not satisfactory. Many users complained on websites and blogs that the iPhone 3G could not run at the full speed of the 3G network. Even in areas covered by 3G, the signal often slipped back to 2G, and Infineon's baseband chip was the culprit.

At that time, a report in Business Week cited sources as saying that the problem was with Infineon's technology, which was "quite new and not extensively tested outside the laboratory environment." It stated that Apple had "overclocked" the Infineon chip, allowing it to demand a stronger 3G signal than it needed, which led to the situation where if too many people in the same area tried to use the iPhone at the same time, it would switch back to a slower network.

This incident not only affected the reputation of the iPhone but also AT&T, the exclusive seller of Apple mobile phones. Gradually, there was a rift between the three parties, and it turned out to be just because of a small baseband chip.

It should be noted that one of the main reasons why Apple insisted on using Infineon instead of a domestic American manufacturer like Qualcomm on the first-generation iPhone was cost. According to the teardown report, the material cost of the Infineon chip was $15.25, accounting for 6.1% of the total cost of the iPhone. At the same time, the BlackBerry Storm equipped with Qualcomm's MSM7600 baseband processor had a cost of about $35, accounting for 17.2% of the total cost. The cost was more than double, so it's no wonder that Apple insisted on using Infineon's baseband on the iPhone even though it knew Infineon was not very good.

However, it is somewhat ironic that, in addition to saving some money on chip costs, Apple also paid a lot of tuition fees: According to documents submitted by Apple, since the release of the iPhone in 2007, Apple has been indirectly paying patent fees to Qualcomm. In 2007, Apple released the first iPhone using Infineon's baseband processor chipset, but due to patent issues, it still needed to pay patent fees to Qualcomm. At that time, Qualcomm did not grant a license to Apple directly according to the customary "FRAND" terms, but signed a confidential patent license agreement with specific Apple contract manufacturers ("CMs"), that is, third-party manufacturers who manufacture and assemble Apple's products, and the CMs paid Qualcomm's patent usage fees, ultimately passing all the costs to Apple.

In order to reduce the excessively high patent fees, Apple signed a "marketing incentive agreement" with Qualcomm in 2007, prohibiting Apple from selling WiMAX terminals, which is an emerging 4G standard and has a competitive relationship with LTE. Qualcomm lacks meaningful patents in this area. One party is to reduce costs, and the other party is to invest in the future market. In this way, a rare consensus was reached.But these agreements did not solve the network issues of the iPhone; instead, they pushed the problem back. The Infineon baseband only supports 3GSM (WCDMA), which means it can only support AT&T in the United States, and after jailbreaking, it can only additionally support T-Mobile. When the largest operator in the United States, Verizon, extended an olive branch to the popular iPhone, Apple ultimately compromised for the money.

In June 2010, the iPhone 4 was released. Although it still only supported 3GSM networks, rumors about the launch of a CDMA version of the iPhone were rampant before its release. With the official launch of the Verizon iPhone in January 2011, AT&T and Infineon's monopoly on GSM iPhones became a thing of the past. Qualcomm, as one of the largest holders of CDMA patents, became Apple's new supplier without any surprises.

According to the teardown, the Verizon version of the iPhone 4 is equipped with Qualcomm's MDM6600 baseband processor, which not only supports the CDMA standard but also supports the GSM standard (hardware only), including up to 14.4 Mbps 3G HSPA+ and up to 14.7 Mbps EV-DO rev B... In addition, the "death grip" issue of the Verizon version of the iPhone 4 was greatly reduced, and the mobile hotspot function was also added.

To borrow a slogan from Apple: The iPhone with Qualcomm's baseband, the only difference is that everything is different.

Betraying Qualcomm and going its own way.

In the early years, the Qualcomm processor integrated with the baseband was called "buying a baseband and getting a processor for free" by many people. Although it was just a joke, it also revealed the strength of Qualcomm's baseband.

Apple obviously tasted the sweetness and, after experiencing the power of Qualcomm's baseband on the iPhone 4, completely switched to Qualcomm for the next generation of the iPhone 4s, and Apple's history with Infineon's baseband came to an end.

In order to retain a large customer like Apple, Qualcomm also made a small sacrifice: in 2011, Qualcomm and Apple signed an agreement to supply exclusively in exchange for a patent fee discount. From this year on, Apple obtained an annual fee of 1 billion US dollars from Qualcomm by ensuring the exclusive position of Qualcomm's baseband and not cooperating with regulatory authorities to investigate Qualcomm. In simple terms, it was a kickback.

This kickback was very advantageous in the early days because Qualcomm's patent fees were calculated based on the number of mobile phones shipped, charging 5% of the selling price for patent authorization fees per phone, which included communication patents from 2G to 4G. Before 2013, the iPhone's shipment volume just exceeded 100 million, and the starting price was still at 599 US dollars. After deducting the rebate given by Qualcomm, it did not increase too much on the previous fees.

However, as the annual shipment volume of the iPhone broke through 200 million, and higher-priced iPhone models were introduced, including Cook and other executives could not sit still. The 1 billion US dollars from Qualcomm every year was fixed, but the money Apple had to pay to Qualcomm was rising. In the end, it meant that Apple had to pay an additional several billion US dollars every year.For Apple, the baseband is just a small chip, not worth so much money. In an interview with Bloomberg Businessweek, an Apple executive revealed the company's thoughts. In his view, the mobile modem is just one of many components and has no special significance. He pointed out that if the cellular network fails, you can use Wi-Fi to access the internet, which uses a different chip. In addition, phones are no longer just phones; they are also navigation tools, digital wallets, health monitors, cameras, and more. All these functions can work with or without mobile service.

"Cellular connectivity is important," said the Apple executive, "but it's not as important as it used to be." Behind him, another Apple representative displayed two versions of the iPhone 7 on a table: one with 128 GB of memory, which Apple sells for $750. The other has a capacity of 256 GB and is priced $100 higher. He asked, considering these two devices are the same in other aspects, is it fair for Qualcomm to charge up to $5 for the technology in the more expensive phone?

Although the rebate agreement between Apple and Qualcomm was until 2016, Apple had already started to break free from Qualcomm's dominance earlier. Not only did they cooperate with Intel on basebands, but they also urged Samsung to pressure the South Korean antitrust regulator to strengthen its investigation into Qualcomm since 2014.

It is worth mentioning that Intel did not originally have a baseband business. This business was acquired for $1.4 billion from Infineon after 2010. Apple, Qualcomm, Intel, Samsung, South Korea... this exciting political and business drama unfolded rapidly.

In September 2016, Apple released the iPhone 7, officially introducing Intel basebands. Qualcomm stopped the annual $1 billion rebate on this basis, and Apple correspondingly stopped paying the patent license fee that should have been paid to Qualcomm.

At the end of 2016, the South Korean regulatory agency fined Qualcomm $850 million for "abusing market dominance" and announced that it would order Qualcomm to change its pricing plan. Three weeks later, the US Federal Trade Commission accused Qualcomm of adopting anti-competitive strategies.

In just a few weeks, Qualcomm's market value, which was over $100 billion in December 2016, fell by a quarter. In this baseband war without smoke, Apple, with its vast resources, seemed to have already won.

So, has Apple achieved its wish?

Since 2017, Apple and Qualcomm have launched a patent war, doing their utmost to block each other all over the world. Both sides believe they are the righteous ones. Apple's lawyers shouted in court that Qualcomm's fees are unreasonable, and they want fairness. Qualcomm's lawyers said that Apple's tricks have been seen through, and they will not succumb to such coercion.To be honest, it's really a waste that Netflix hasn't turned the peak confrontation between the two companies into a TV series.

Especially in the final lawsuit, Apple's lawyer used a KFC restaurant as a metaphor for Qualcomm's charging policy: You have to go to the counter to buy a food license first, and then you can buy a piece of fried chicken to enjoy.

Qualcomm, on the other hand, said that the fried chicken is the baseband you bought, the patent fee is the fries, and what Apple paid for is a meal, not a piece of fried chicken.

In the end, Qualcomm and Apple reached a settlement after the lawsuit in April 2019. The settlement agreement included an undisclosed amount of money paid by Apple to Qualcomm and a chip supply agreement, which is valid for 6 years and can be extended for another two years.

It's not hard to guess the reason for Qualcomm's compromise. As long as the popular iPhone is still equipped with Qualcomm's baseband, the chip cost and patent licensing fee will continue to flow in.

The reason for Apple's compromise, after all twists and turns, has returned to the starting point. Intel's baseband comes from Infineon, and Intel itself does not have a deep foundation in the field of communication. Apple also does not have it. The baseband developed by the two companies naturally inherits all the previous shortcomings. According to the test results of foreign media, on the most common LTE frequency bands in the United States, the iPhone X with Qualcomm's baseband has always performed better in LTE speed than the Intel baseband, even though Apple deliberately restricted the Qualcomm baseband.

This gap may not be very obvious in the same model comparison, but once it is put into many Android flagship machines, the iPhone's shortcoming is immediately exposed. From the iPhone 7 to the iPhone 11, these four generations of models have almost swept the first place in the worst signal list.

Seeing that 5G is coming, Apple and Intel have not yet figured out the 4G baseband. It doesn't make much sense to hold on, and they can only rush to cooperate with Qualcomm again.

Of course, Apple has never stopped the ambition to kick out Qualcomm and make more money. In July 2019, shortly after Intel announced its exit from the baseband business, Apple announced that it would take over with 1 billion US dollars and start its own baseband.

The mountain that cannot be crossedQualcomm's headquarters is located in an office park in the suburbs, and the most eye-catching feature is the patent wall. This two-story high installation displays hundreds of Qualcomm's patent applications. The original CDMA patent is enlarged and annotated, and the original CDMA technical specifications are also printed out and preserved behind glass in a separate company museum nearby, spanning 685 pages. The CDMA specifications were approved by the wireless industry in 1993 and incorporated into the 3G specifications, taking nearly a decade to gain popularity.

In Bloomberg's interview report on Qualcomm, top engineer Grob stands in front of a mechanical arm, the head of which is a human model, and the hand is usually on the head of this human model. This is one of the dozens of laboratories at Qualcomm's headquarters in San Diego, part of a research and development business that costs more than 5 billion US dollars annually. When asked about the "Frankenstein" nature of the laboratory, Grob said indifferently, "There are many heads and body parts here." He explained that each body part is injected with a saline solution to simulate the viscous substance inside a real human body. The purpose of this is to test the impact of the presence of flesh on signal strength and to ensure that Qualcomm's modems can adapt to this impact.

After considering all possible factors that may affect signal strength, wireless companies must also manage the spectrum. Major mobile operators all have some spectrum, and within specific spectrum blocks are channels, each with hundreds of channels, and each mobile phone needs to access at least one channel. The trick of the wireless business lies in how to compress a large amount of mobile phone data into a channel, and then transmit these signals between the mobile phone and the radio tower, even if the signals are reflected by buildings, moving at high speed, or obscured by leaves.

It is clear that Qualcomm's baseband is not just a simple silicon chip. Just like Apple's self-developed A and M series chips, the technology and manpower invested behind it have already accumulated to an exaggerated extent. Even if we put aside the limitations of intellectual property rights, it is by no means easy for Apple to make a baseband that can match Qualcomm's.

As for patents, unless Apple can build a time machine to go back to the 1990s, it is impossible to bypass CDMA patents. This means that even if the chip is made, it may still need to continue paying Qualcomm, and the only thing that changes is the logo on the chip.

The input-output ratio is also an important aspect. Spending tens of billions of dollars to save a few billion dollars annually, even Apple, which has hundreds of billions of dollars, cannot make this muddled account.

Apple has been making basebands since 2021, and apart from the speculation of analysts, there has been no specific result so far. The reasons are nothing more than these three points. The baseband manufacturers have been sifted through the waves, and now only Qualcomm, Huawei, Samsung, MediaTek, and Unisoc are left, and the fierce competition behind the technology and patents can be imagined.

At the launch of the new iPhone, Cook and executives are full of enthusiasm for new materials, new processes, and the seamless integration of software and hardware. It seems that it is Apple's innovation that has given all this, even when boasting about the 5G network.

However, Apple cannot manipulate everything. At least, it has not yet mastered the baseband.