Network Security Internet Technology Development Database Servers Mobile Phone Android Software Apple Software Computer Software News IT Information

In addition to Weibo, there is also WeChat

Please pay attention

WeChat public account

Shulou

Chip Wall Breaker (5): the Mobile Age discovered by Acorn and ARM

2025-03-28 Update From: SLTechnology News&Howtos shulou NAV: SLTechnology News&Howtos > Internet Technology >

Share

Shulou(Shulou.com)06/02 Report--

If you were a young, ambitious entrepreneur who had to start from scratch and had a chance to go back to any period of the IT industry from scratch, which one would you choose? I think you will choose the exciting '70s.

If the 1950s were the era of the birth of semiconductor technology and the boom of tube computers, and the 1960s were the era of the beginning of the semiconductor industry and the rise of mainframes in the commercial market, then the 1970s will usher in the dawn of the consumer market of the IT industry. At this time, several companies are still in charge of the world territory of IT.

Paul Allen and Bill Gates started their business in 1975 and founded Micro-Soft the following year. In 1978, it was officially renamed Microsoft, which is now Microsoft.

On April 1, 1976, Steve Jobs and his childhood friend Steve Wozniak co-founded Apple computer in the garage of Jobs' parents' home in California.

In 1979, Herman Hauser, who just graduated from Cambridge University in England, founded Acorn computer Company with Chris Kerry.

(Acorn Computer Ltd logo)

There is no need to say much about the position of Microsoft and Apple today, but many people are not familiar with this British-made Acorn computer. If we say that Acorn computers are the forerunners of ARM, which today provides the system architecture for billions of mobile terminal chips, you may have a sense of enlightenment.

We know that Apple ushered in the era of personal computers, but because of Master Joe's insistence on going his own way, he was defeated by IBM, Compaq, Dell and other companies that took the route of Wintel alliance. Apple's Mac is still a minority representative of the PC era. The Acorn computer has not been able to live through its glory days, but it has left ARM for the world as the cornerstone of the surging computing power that underpins Apple, Android and other smart mobile terminals.

In June 2020, Apple announced on WWDC 2020 that it would install the self-developed chip with ARM instruction set on Mac products, so as to completely open the underlying computing architecture of mobile and PC. Apple's choice seems to be a perfect reincarnation. You know, ARM was founded by Acorn computer, Apple and VLSI, a semiconductor contract manufacturer.

In this issue, we will return to the historical scene where Acorn was born and take a fresh look at the karma that led to ARM and the mobile era ushered in by the ARM processor architecture.

A large number of wizards: the Acorn computer that created ARM

Back in the early 1970s, when the computer was just an expensive device used by large enterprises and government agencies, inventing and making personal computers was naturally the most fashionable thing for electronic enthusiasts at that time.

When Jobs saw the talent of Woz, who is a few years older than him, in making computers, he seemed to have decided on the path of personal computers. Finally, in 1976, they made a prototype of Apple 1 and successfully sold the first batch of Apple computers.

At first, Jobs insisted on using Apple as the company's name. it is said that he mainly ate Apple during his two years of penance in India, which became the source of his strength, and that Apple would be at the top of the phone yellow pages in the future.

This "smart" idea also seemed to inspire the founders of Acorn, Hauser and Kerry, who named the company Acorn in 1979, mainly in the hope that their PC business would grow into a vibrant oak tree like acorn. At the same time, they also want to rank ahead of Apple in the phone yellow pages.

Who would have thought that such two companies named after "fruit" could become giants in the field of intelligent mobile terminals and mobile terminal processors, respectively.

At that time, it was just the beginning. Whether it is a technical enthusiast who fiddles with electronic devices all day in the garage, or a PhD in physics trained in Cambridge, England, and a marketing genius with little technical background, they can even get involved in personal computers, which represents the promising direction of the IT industry at that time, which shows the vitality and recklessness at the beginning of an industry.

For Acorn, in addition to the two ambitious founders, the next two characters are even more crucial. One is Sophie Wilson, the first technician of Acorn, who had just graduated from the Department of Mathematics at Cambridge University when she joined Acorn. To put it simply, she was the earliest developer of the Acorn computer and the first developer of the ARM architecture instruction set. The other is legendary engineer Steve Fobar, who worked with Wilson for a week to come up with a prototype of a microcomputer that satisfied BBC, and then he worked with Wilson on the development of the ARM processor, and he was in charge of the chip design.

It is the times that make the genius, and the genius also makes the times.

Acorn's first generation of cheap personal computers, the Acorn Atom, helped it initially open up the home market. What really gave Acorn a foothold in the microcomputer market was the £1.3 million order from BBC Micro in 1981. This project is inseparable from the active efforts of Chris, Herman in the internal successful mediation, successfully stimulated the potential of Wilson and Fobar. Before the BBC team came to visit again, the two men used the Acorn Proton version, which was still in development at the time, as a prototype, and it took five days and five nights to build a BBC Micro physical prototype. Since then, BBC Micro has been a great success, selling a total of 1.5 million units and won the Queen's Technical Award (Queen's Award of Technology) in 1984.

(photo of project participants at BBC Micro commemoration in 2008)

The opportunity to promote Acorn's own processor development is that after IBM launched the second generation of microcomputers for the commercial market in 1983, Acorn also intends to enter the commercial market of microcomputers. However, Mostec's 2MHz 6502 processor, which was originally used in BBC Micro, can not meet the new hardware requirements, and there is no processor on the market to satisfy Acorn.

It is said that Wilson and Fobar looked for all the chips that might be used at the time and thought Intel's 286 chip was not bad. When they offered to cooperate with Intel to get the license for the 286 chip, they were flatly rejected by Intel. The price of this refusal is that Intel has trained a nightmare rival for the new century.

The balance of opportunity tilts just towards Acorn. At that time, the University of California, Berkeley, under the chairmanship of Professor David Patterson, proposed the Berkeley reduced instruction set Program (Berkeley RISC Plan) white paper. It was inspired by this that Wilson and Fobar began to develop 32-bit microprocessor chips based on RISC for the new generation of BBC microcomputers.

(Sophie Wilson is introducing the ARM development process)

In October 1983, the ARM program was officially launched. Following the streamlining of the RISC architecture, Wilson soon wrote the first prototype of ARM in BBC Basic. After 18 months of research and testing, in April 1985, VLSI, the chip contract manufacturer of Acorn, produced the first ARM chip using RISC instruction set, where the full name of ARM is Acorn RISC Machine.

(ARM-1 chip on BBC Micro)

At 1: 00 p. M. on April 26th, 1985, the first batch of ARM chips came back from VLSI, was put directly into the development system, and started after one or two adjustments. At 3 p.m., the screen says "Hello World,I am Arm". ARM-1, the world's first commercial RISC processor, has been successfully run in Acorn computer Company.

However, at this time, the Acorn computer company lost a lot of money as a result of business mistakes, and Acorn had to transfer nearly half of its stake to Italy's Olivetti computer company at a low price of 12 million pounds in 1985 to repay the debt. After the successful development of the ARM chip, Acorn uses the ARM development system and an Acorn Archimedes system to sell commercial products.

Since then, the development of Acorn computers has been calm. Because in 1990, ARM (full name is Advanced RISC Machines Ltd), co-founded by Acorn computer, Apple and VLSI, the contract manufacturer, was founded.

The new historical process and glory will belong to ARM.

RISC's Spark: ARM chose the right track

Before continuing with the story of ARM, it is necessary to talk briefly about the obscure field of instruction set.

In 1961, IBM, with the support of Watson Jr., was led by Vice President Vincent Lilsson and was prepared to invest $5 billion in the development of IBM360 computers. In 1964, IBM series computers were successfully developed and became epoch-making products.

In the process of development, IBM has overcome a series of software and hardware difficulties, such as computer instruction set, integrated circuit, compatible operating system, database and so on. Among them, System-360 system, a new integrated general instruction set architecture (Universal Instruction Set Architecture), has become the first commercial instruction set architecture in the history of computer development.

(IBM System360 mainframe)

It can be said that the essence of instruction set is a set of "standard language" for communication between hardware and software code. The processor is the core hardware and the operating system is the basic software. The software running in the computer must be developed based on the architecture of the instruction set in order to achieve the effect of normal operation. This mode of building from processor to operating system to basic application software is a compatible ecology of a "standard language" based on instruction set architecture.

We often talk about processor architecture, that is, processors based on instruction set architecture. Once the instruction set is bound with the operating system combination, it will form an insurmountable "ecological wall" for other players. We now know that the most unbreakable "ecological wall" leading the PC era is the Wintel alliance built by Intel x86 processors and Microsoft's Windows operating system. The "ecological wall" in the era of mobile Internet is the ARM alliance formed by the mobile processor ecology formed by ARM's RISC instruction set architecture and Android and iOS operating systems.

During the development of instruction set system, there are two different optimization directions: complex instruction set computing (Complex Instruction Set Computing,CISC) and reduced instruction set computing (Reduced Instruction Set Computing,RISC).

CISC is to improve the execution speed of the computer by setting some instructions with complex functions and changing some of the commonly used functions realized by software to the instruction system of hardware, which is characterized by complex design, high power consumption and different execution time of instructions, and its advantage lies in high computing performance. The design idea of RISC is to simplify the computer instruction function as far as possible, only those instructions that are simple and can be executed in one beat, and the more complex functions are realized by a subroutine, which is characterized by the same execution time of each instruction, which can reduce the average execution cycle of instructions and improve the speed of main frequency. Another advantage is that the power consumption is very low, but it is not as good as CISC in general purpose calculation.

(famous Intel 8086 processor)

Although in the 1980s, academia thought that CISC was out of date. However, because Intel did not have the RISC instruction set when developing the 8086 processor, it adopted the design of CISC. Since then, Intel's processor series have adopted the CISC instruction set.

Since then, due to considerable revenue, Intel can continue to invest in the research and development of CISC processors, ensuring that its processor performance continues to exceed that of RISC processors, and finally won the dominant position in PC computer processors.

In the 1980s, Motorola, IBM, SUN, SGI, DEC and HP were all producing their own PC processors in the RISC processor market, but due to competition with each other and a big price war, they were unable to compete with Intel's CISC processors and changed sides one after another. Fortunately, the glimmer of RISC is still preserved in ARM.

(in 1990, ARM's start-up office.)

ARM, a new company founded in 1990, has only 12 engineers involved, not including Wilson and Fobar, the founders of the company. The office is just a barn in Cambridge. Acorn and Apple each own 43% of the shares, while VLSI holds the rest of the shares and becomes the semiconductor contract manufacturer of ARM and the first chip manufacturer to be licensed by ARM.

(Apple Newton PDA with ARM-6 processor)

In 1993, ARM once again encountered a cold shoulder in the market when he partnered with Apple to develop an Apple Newton MessagePad with an ARM processor. ARM realizes that the success of an enterprise cannot depend on individual products.

At that time, ARM creatively launched an IP-licensed business model from Motorola's favorite new CEO, Rubin Saxby. Since then, chip manufacturers can obtain the IP copyright of RISC instruction set processors and most of the revenue as long as they obtain the license of ARM processors from ARM and pay the pre-license fee and the patent fee for the later production of the chip. On the other hand, ARM does not have to bear the operational risk caused by product development failure or poor sales.

This is a mutually beneficial and win-win business innovation, and it also lays a foundation for the high division of labor and accelerated development of the RISC chip industry in the future.

In 1993, ARM's cooperation with Texas Instruments, Samsung, Sharp and other semiconductor giants established a reputation for the promotion of ARM and confirmed the feasibility of the business model authorized by IP.

(Motorola 6110 is equipped with an ARM-7 processor)

At this time, the development of ARM also caught up with the "right time and right place" of the mobile device revolution. At that time, the Nokia 6110 became the first GSM phone with an ARM processor. In order to meet Nokia's memory reduction requirements, ARM specially developed a 16-bit custom instruction set, which significantly reduced memory, and was eventually produced and sold to Nokia by Texas Instruments. The listing of 6110 has been a great success. Qualcomm, Freescale, DEC have joined the ARM-7 licensing camp, and later licensed to more than 170companies, so far produced a total of more than 10 billion chips. ARM-7 has also become ARM's flagship series of mobile processors.

In 1998, ARM listed in both the United States and the United Kingdom. In just a few months, ARM has become a public company valued at more than $1 billion, and today it is a proper unicorn, but this is only the beginning of ARM's take-off.

Take advantage of potential growth: ARM ushered in the Mobile era

Although in 2001, the stock market of the Internet industry market collapsed, the stock price of the whole industry fell sharply, and ARM revenue also plummeted, ARM did not suffer a great deal of energy. That year, Warren East succeeded Saxby and became CEO,ARM, which is controlled by ARM, and began to implement a roadmap for the next five years. The goal of ARM becoming a RISC processor architecture standard is being realized.

(ARM Cortex-A15 chip)

After the launch of ARM, there is plenty of money to start the design of higher-end processors. In 2001, ARM-926EJ-S launched the IP, which was licensed to more than 100 companies and shipped 5 billion units. Since then, ARM has launched ARM-10, ARM-11 series architecture one after another. In 2005, I realized that in addition to the high-end processor market like ARM-9, there is also a market for low-cost and low-power microcontrollers. Therefore, ARM proposes the Cortex product line and divides the processor architecture into:

Cortex-A for high performance, Cortex-R for real-time control scenarios, and Cortex-M for microcontroller market.

Since then, with the hot sales of iPhone smartphones and the launch of Android, the world has entered the era of smart mobile phones, and ARM is about to ascend to the throne of the mobile terminal era with the support of Apple and a number of Android manufacturers and mobile chip manufacturers. Because ARM processors account for 90 per cent of the mobile terminal market, ARM's annual shipments have risen from 10 billion in 2013 to 20 billion now.

(cooperative Ecology of ARM)

At this point, ARM has finally embarked on a completely different business model and market positioning from Intel. It seems that the two should have developed in their respective market areas, so that "well water does not invade river water". But with the saturation of the PC computer market and the arrival of the era of mobile Internet, Intel has already focused on mobile devices, and ARM has also extended its processors to the server market dominated by Intel.

Intel bought the StrongARM license from DEC as early as 1997 and upgraded StrongARM to Xscale. However, Intel does not adopt a highly integrated design pattern for XScale, resulting in strong performance but high power consumption. In 2005, Intel turned down an order for processor design from Apple, and the following year, due to business contraction, the sale of XScale was tantamount to ceding the opportunity to occupy mobile processors and missing the coming mobile era.

When Intel realized the threat of ARM again and wanted to use the embedded Soc--Atom designed by x86 to face the ARM ecology, it could only be a complete failure.

The year of renewed war was in 2011, when Microsoft announced that it would officially support ARM processors in the next version of Windows, which began to shake Intel's x86 processor market position. By 2016, Intel shut down its Atom production line because of huge losses, while historical shipments of ARM chips reached a staggering 100 billion. Of course, the ARM ecosystem has repeatedly failed in the attack on the server market occupied by x86 systems.

(ARM Neoverse chip for high-performance server applications)

Now, in the mobile terminal market, the huge Intel is difficult to pose any threat to ARM. As the boundaries between mobile devices and PC devices become increasingly blurred, and data centers are looking for more diverse, higher-performance, low-power servers, ARM once again launches a new charge on PC processor chips, server chips and AI chips.

At the same time, with the arrival of the smart era of the Internet of everything, more and more Internet of things devices will be equipped with more than one embedded chip. And that's what ARM is good at, but it's also a new battleground that Intel wants to focus on in the future.

After that, we will see an all-out war in the field that Intel sticks to and in the coming field of smart things in the midst of the rapid development of ARM ecology.

ARM's successful "harmony of time, place and people"

The reason why ARM is able to achieve today's market position and business success is mainly due to its innovative business model-IP licensing model.

Different from Intel's asset-heavy, high-profit monopoly model, in the field of mobile terminals, low-cost, low-profit processors can not support this heavy investment model. The licensing model initiated by ARM separates the architecture and design of the chip (Fabless, such as Qualcomm, Nvidia, Huawei Hayes) and production (Foudary, such as TSMC), which not only shares the cost, but also improves the production efficiency and the speed of new process iteration, thus forming an increasingly prosperous ARM ecology.

(Apple chip with ARM architecture)

According to the different needs and capabilities of chip design companies, ARM provides three different modes of external authorization, including:

First, processor authorization. The chip designer that buys IP only needs to produce according to the chip drawings designed by ARM.

Second, processor optimization package and physical IP package authorization. Chip design enterprises can directly get a series of design schemes to complete the chip production, but the degree of freedom is lower, and the processor type, foundry and process are all specified by ARM.

Third, architecture and instruction set authorization. For powerful chip designers, such as Apple, Qualcomm, Samsung and Huawei Hayes, they can buy ARM architecture and instruction set directly and design their own processors compatible with ARM instruction set.

Currently, ARM has more than 1000 processor licensing partners, 320 processor optimization packages and physical IP package licensing partners, and 15 architecture and instruction set licensing companies worldwide. With the support of many authorized companies, ARM processors shipped 15 billion units in 2015, with a total of more than 100 billion units shipped in history.

ARM chooses up to three partners each time it develops a next-generation processor, IP. These selected companies will be able to understand ARM's design earlier and take the lead in new product development, but they will also help ARM debug, test, and provide feedback to ARM, so ARM can ensure smooth development and speed up application.

In the profit model, the profit of ARM depends entirely on IP authorization, and the profit depends entirely on the number of chips that can be sold by licensors, partners and customers, which is closely bound with the enterprises that design, produce and sell chips, and work together to maximize the benefits of products and achieve win-win results.

In the choice of processor development direction, the RISC instruction set processor selected by ARM was temporarily defeated in the confrontation with Intel's CISC instruction set, but there is no doubt that the trend of low power consumption represented by RISC is more in line with the development direction of the mobile era.

Now, this advantage of the RISC instruction set is more obvious. In other words, it is a more advantageous process for processors to strive to break through the limit of performance in the case of low power consumption than to reduce energy consumption while maintaining performance improvement. The expansion of ARM's processor chips into Intel's processor chips is the best footnote to this trend.

To sum up, we can see that the outbreak of the mobile era represents the right time, the location represented by the advanced performance of RISC and the people represented by the innovative business model of ARM, together constitute the prosperous ecology and global career map of today's ARM.

As for the cooperation and entanglement between ARM and China's chip industry, we will discuss it in another chapter.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.

Views: 0

*The comments in the above article only represent the author's personal views and do not represent the views and positions of this website. If you have more insights, please feel free to contribute and share.

Share To

Internet Technology

Wechat

© 2024 shulou.com SLNews company. All rights reserved.

12
Report