Science Technology

Apple is becoming a leading chip company

Li Shi’s introduction: Over the years, Apple’s products have become more and more prominent in the chip field, which gradually makes the public realize that Apple has formed a wide moat in the chip field.
Apple is becoming a leading chip company
Li Ping | Author
Apple’s advantages in the field of chips are increasingly prominent.
On March 9, Beijing time, Apple held a new product launch in the spring of 2022. In addition to five new products, including the new color matching iPhone 13 and iPhone 13 Pro, the new iPhone SE, and Mac Studio and Studio Display, the M1 Ultra, known as the world’s strongest desktop processor, also received global attention.
Apple is becoming a leading chip company
As the fourth generation product of the M1 series, the M1 Ultra uses the new UltraFusion packaging architecture to interconnect two M1 Max chips, thus achieving higher performance and lower energy consumption. It is reported that the M1 Ultra contains up to 114 billion transistors, supports up to 128GB of unified memory, and has a 20-core CPU and a 64-core GPU, which is 8 times faster than the M1. In addition, under high performance, the performance power consumption of the M1 Ultra is 90% lower than that of today’s 16-core computers.
The release of M1 Ultra not only enriches Apple’s M1 processor camp, but also makes it easier for Apple to say goodbye to its “old friend” Intel. An increasingly powerful Apple chip empire is also changing the competitive landscape of the global semiconductor industry.
Apple was also controlled by others
Apple’s core making plan began in the 1990s.
In 1997, after Jobs, the founder of Apple who had been expelled before, was invited back to the company, he found that the personal computer market at that time had been dominated by the Wintel (Windows&Intel) alliance, and Apple’s market share was poor. In the face of the combination of software and hardware between Microsoft and Intel, Jobs, together with IBM and Motorola, established the AMI PowerPC (Apple, IBM, Motorola) alliance to jointly develop and launch the PowerPC chip in order to regain the dominance of the PC market.
According to the original idea of the AMI alliance, PowerPC chips are mainly developed by IBM and Motorola, while Apple is responsible for assembling PowerPC chips on personal computers. In addition, the chips are also provided to companies outside the alliance.
PowerPC chip adopts the simplified RISC chip architecture of IBM Power, inherits the powerful computing performance of IBM high-end server workstation and Motorola’s top chip technology reserves, and has the advantages of good scalability, convenience and flexibility. However, PowerPC chips also have disadvantages such as high energy consumption and high price.
Finally, although Apple Mac with PowerPC chip was snapped up by Apple fans, it did not continue to expand its market share.
In fact, the proportion of other manufacturers using PowerPC chips is very low except Apple Computer. This creates a contradiction between R&D investment and sales profit.
For IBM, supplying chips to Apple requires a huge investment in the development of chipsets, compilers and other supporting technologies. However, since Apple was a niche market at that time, IBM’s market share of PC chips was very limited and it was difficult to make money. Apple, on the other hand, publicly criticized IBM for its problems in supply, and believed that the power Mac G5 chip produced by IBM could not be used in notebook computers due to heat dissipation and power consumption problems. The two sides were in constant disagreement.
Although it is called joint research and development, the research and development of PowerPC chips mainly depends on IBM, not Apple. Due to the disproportionate input and output, IBM and Motorola had to turn part of their design efforts to the embedded PowerPC chips in the field of automobile manufacturing, rather than focusing on meeting the needs of Apple Mac. Jobs himself complained about this, but there was nothing he could do.
In 2003, when Jobs launched the Mac with PowerPC processor, he publicly said that the frequency of the processor would reach 3GHz within 12 months. But in fact, after 24 months, the 3GHz processor still disappeared, which also gave Jobs the first taste of key components being controlled by others.
In 2005, after losing hundreds of millions of dollars on PowerPC chips, Apple decided to withdraw from the AMI alliance and switch to Intel. On June 6, 2006, Apple officially announced that it would use Intel processors in Macintosh computers and abandon the PowerPC architecture, ending its decade-long partnership with IBM. In response, Jobs said, “Intel will develop the most powerful processor products in the future. I believe that Intel technology will help us build the best personal computers in the next decade.”
Apple is becoming a leading chip company
However, the honeymoon period between the two sides did not last long, and the cooperation soon broke. On the one hand, it is difficult for Intel to keep pace with the fast-paced Mac, and the release plan of Apple’s new products is often subject to Intel chip delay and security issues. On the other hand, perhaps in the era of PowerPC chips, the lessons learned from others are too profound. Apple has never given up its efforts to achieve “chip independence”. Therefore, although Apple agreed to transfer its Mac platform to Intel’s x86, it has made plans internally that other products do not involve Intel.
In fact, the cooperation between Apple and Intel has been full of contradictions. As early as 1997, when Jobs met with Andy Grove, the then chief operating officer of Intel, he asked to enjoy the “maximum customer price”, which was rejected by Grove, so Apple only cooperated with Intel for one year at that time.
In response, Grove once said: “Since then, Intel has been worthless in Jobs’ eyes. No matter what we do, we can’t change his mind.”
According to Paul Otellini, the former CEO of Intel, Intel had the opportunity to become the processor supplier of Apple’s first generation iPhone. Apple expressed strong interest in a chip of Intel at that time, but the price proposed was far lower than expected, so Intel lost the order. Finally, Apple chose the ARM architecture chip from Samsung.
It is not difficult to see that, as the leader of their respective fields, the cooperation between Apple and Intel is full of bargaining contradictions. For Apple, which attaches great importance to supply chain security, Intel, which is in a monopoly position, has always been an unsafe factor. After all, for Apple, which adheres to the integration of hardware and software, chips are the core hardware, and self-research is also a must.
Finally, when Apple revived with the popularity of iPhone, iPad and other products, it immediately opened the road of self-development of mobile phone chips. At the same time, Intel lost the order of Apple’s mobile phone due to the price problem, and completely lost the opportunity to change the world.
Self-research road of mobile phone chip
Although the PowerPC processor exited with failure, it really kicked off Apple’s self-developed chip. While fully recognizing the importance of the chip, Jobs also understood the difficulty of the chip’s self-development. For this reason, Apple has chosen to start with the mobile chip which is still in the early stage of the industry development, and continue to cooperate with Intel in the field of desktop chip.
In 2010, Apple launched its first self-developed mobile phone chip, Apple A4, with the main frequency exceeding the 650MHz of the public version of ARM at that time and breaking through 1GHz at one stroke. Later, this chip was successively carried on the first generation of iPad, iPhone 4 and iPod touch 4.
Facts have proved that the iPhone 4 with A4 chip has become a milestone in the development of Apple’s smart phone, and the iPad has almost become the synonym of tablet computer. Since then, A-series chips have always been the representative of high-performance chips on mobile platforms, with generations of improvements in process, CPU architecture and GPU core.
After gaining a foothold in mobile chip, Apple’s chip research and development has also expanded from mobile tablet to smart watch, true wireless headset and other categories, such as S series chips on Apple Watch and W series wireless Bluetooth chips.
Apple is becoming a leading chip company
In terms of chip architecture, Apple A series, S series and W series chips all adopt ARM architecture. Compared with Intel’s x86 architecture, ARM has the advantages of low energy consumption, low cost and high performance, which has a significant impact on consumer electronics products. In particular, the longer battery life, thinner and fanless design and other functions make the ARM based Apple chip obtain incomparable advantages in system stability, endurance and other aspects.
More important than performance improvement, Apple has greatly built up the wall of Apple’s closed ecosystem through chip self-development, enabling the Apple ecosystem to more smoothly realize internal circulation.
Self-developed chips quickly brought positive incentives to Apple. Because each generation of A series has obvious performance advantages compared with other mobile phone chips in the same period, each generation of iPhone has a more advanced comprehensive experience compared with the flagship in the same period – although Apple is selling more and more expensive, it is selling better and better.
In 2017, the launch of Apple’s A11 Bonic bionic chip ushered in the AI era of smart phones. The addition of neural engines further improved the all-round functions and experience of mobile phones through algorithms, such as AR, face recognition, and image integration. At the same time, A11 is also the most independent generation of A series processors in Apple’s history, including self-developed CPU, self-developed GPU, self-developed ISP, self-developed decoder, etc., especially the first time it uses the self-designed GPU core.
However, due to the lack of technology accumulation in the communication field and the lack of patents, Apple still needs to purchase baseband chips at this time. In the field of baseband chips, Qualcomm’s technical strength and market share are in the leading position, but there are huge patent licensing lawsuits with Apple. For this reason, Apple chose to purchase baseband processors from Intel and Qualcomm at the same time.
Intel has once again become Apple’s “pig teammate”. Later, consumers reported that the Internet performance of the iPhone 8 with Intel processor was significantly weaker than that of the Qualcomm version, so Apple could only reduce the Internet speed of all mobile phones.
What’s more, in 2019, because Intel’s 5G baseband chip was difficult to produce, the release of Apple’s 5G iPhone was forced to be delayed by one year, and Apple once again tasted the taste of core technology being controlled by others.
In the end, Apple had to shake hands with Qualcomm and completely return to Qualcomm Snapdragon baseband from the generation of iPhone 12, and completely abandoned Intel.
The loss of Apple, a major customer, also put Intel’s baseband business into a complete predicament. Finally, Intel sold the baseband business of mobile phones to Apple at a price of $1 billion, which further increased the self-research ratio of mobile phone chips.
Up to now, Apple’s iPhone 13 series still uses Qualcomm’s Snapdragon X60 5G modem. According to the latest news, Apple plans to let TSMC manufacture 5G iPhone baseband from 2023, and is expected to make an official debut in the iPhone 15 series. It is said that Apple hopes to launch a “high-end baseband”, whose performance will be far superior to Qualcomm products, so the research and development cycle is long.
It is widely believed in the industry that with Apple’s excellent vertical integration capability and product-side software and hardware optimization capability, it has the conditions and strength to improve and optimize Intel baseband chip business, solve signal problems and other technical weaknesses, and then integrate A-series processors and baseband processors to make them perfectly applicable to Apple products and get rid of Qualcomm’s constraints. This will further strengthen Apple’s dominance in the supply chain.
Say goodbye to Intel
Perhaps it is the 5G baseband chip problem that makes Apple completely despair of Intel. At the WWDC Developers Conference in June 2020, Cook publicly announced that in the future, Mac computers will give up Intel processors, and all Apple Mac products will use Apple’s self-developed chips “within two years”.
Five months later, Apple officially launched the first computer chip M1, which was installed on the newly released MacBook Air, MacBook Pro and Mac Mini. This also means that Apple’s laptops and desktops have abandoned the Intel processor that has been used for 15 years.
For Intel, the absence of 5G baseband chips not only hindered its development in the 5G era, but also lost Apple as a major customer in the field of desktop chips. However, the reason why Apple was able to abandon Intel completely was not out of revenge. The fundamental reason was also that the R&D strength of the two companies increased and decreased, and Intel’s chip manufacturing capability lagged behind.
Apple is becoming a leading chip company
For example, as Apple’s first Mac-developed chip, M1 uses the most advanced 5nm technology of TSMC, and contains an amazing 16 billion transistors. It integrates 8-core CPU, 8-core GPU, neural network engine, DRAM memory and other functions on the same chip. The CPU is up to 3.5 times higher and the GPU is up to 2 times higher. Its performance is better than that of Intel processors.
At this time, Intel was still at a standstill in the 14nm process technology. The 10nm process could not be put into production due to yield, capacity bottlenecks and other problems. The 7nm process was delayed again and again, and the nickname “Toothpaste Factory” continued to be established.
After M1, Apple released the more powerful M1 Pro and M1 Max in 2021, and the performance and power consumption are optimized multiple times on the basis of M1. With the support of M1 series chips, the sales of Apple’s Mac products have risen steadily, with the revenue of the fourth quarter of 2021 breaking the $10 billion mark for the first time, setting a record.
Apple is becoming a leading chip company
The chip’s self-development has brought a significant boost to Apple’s financial data. In the fourth quarter of 2021, the overall gross profit rate of Apple’s hardware business reached 38.4%, a new high in six years, and the comprehensive gross profit rate reached 43.8%, a year-on-year increase of 4 percentage points, a new high in 10 years. In fiscal year 2021, Apple’s net profit reached a record $94.68 billion, and the chip’s self-development was the most important contribution.
In addition to performance improvement, Apple’s self-developed chips also have platform integration considerations. Since the M1 chip also adopts ARM architecture, the apps on iPhone and iPad can be synchronized to Mac for use. Apple has thus broken through the barrier between iOS and MacOS, realizing the unification of hardware, software and other aspects, and the ecological closed-loop has been further developed.
Apple in the Cook era has always been criticized for its lack of innovation and lack of iPhone-style “disruptive products”. But in fact, in the past ten years, Cook led Apple to achieve continuous evolution on the bottom of hardware and strengthen its own ecological advantages through chip self-research. This is not only Apple’s focus on the development strategy of software and hardware integration, but also its accumulated technical strength. This should be a good reference for Chinese enterprises suffering from “lack of core”.