'I am 30 centimeters taller than Andy Grove. But whenever I am with him, I think he is a giant.'
Intel's former CEO Andy Grove died in 2017. This was the description of Clayton Christensen at the time. Christensen was the best seller of the word 'disruptive technology'. The author of the book, Professor of Harvard Business School. He said that he missed Grove most. He was able to understand how complex organizations work and how to make Intel benefit from it.
Intel was founded on July 18, 1968. It was this ability of Grove that enabled him to successfully transform the company's direction in the 1980s. Intel's business shifted from large computer storage chips to microprocessors.
Later, Intel and IBM reached an agreement to pre-install Intel processors on all of IBM's personal computers. Under the auspices of this agreement, Intel's technology began to enter Silicon Valley. Intel inside and Jingle sound became the most memorable advertising slogan in modern times. one.
Intel has been occupying a dominant position in this industry. Now that it has been half a century, no company in the world can surpass Intel to produce better and faster microprocessors. This is a miracle industry, and Intel is at The pinnacle of this industry. People tend to think that innovation is uncertain, especially in industries where innovation is heavily dependent on scientists. However, Intel's development is not ambiguous, it continues to introduce processor updates like watches.
In 1965, Gordon Moore, who had not yet become Intel's co-founder, made a bold prediction of the exponential growth of computing performance: the number of microchip transistors etched into computer microprocessors would be two per After doubling in years, computing performance will also increase. After that, Intel made this seemingly unlikely prediction a reality, and Moore's Law continued for decades.
What is the huge impact of Moore's Law? The computing performance of an iPhone today is many times higher than the entire spacecraft of the NASA Apollo mission to the moon in 1969. Without Moore's Law, Google, Facebook, Uber and Airbnb may be It won't exist, Silicon Valley may be just an ordinary place.
Major mistakes: Passing the iPhone After Intel won Apple's Mac business in 2005, Steve Jobs quickly proposed to let Intel design a chip for Apple's smartphone. Intel certainly wants to dominate this emerging field, but Jobs' offer is lower than the forecast cost, and Intel was wrong at the time. Judging the size of the iPhone market, I did not accept Jobs' offer.
Apple has no choice but to build its own chipset through ARM's licensing technology. Google will soon launch a free operating system Android, and Samsung, Huawei and HTC have adopted this system. Qualcomm, NVIDIA and Texas Instruments Acquired ARM's licensing technology and become the preferred supplier of mobile phone manufacturers for energy-efficient and low-cost computing devices.
These competitors from the United States have their own strengths. Qualcomm focuses on mobile phones, and NVIDIA focuses on video game graphics cards. They all outsource production to third-party manufacturers in Asia. But Intel microprocessors sell for about $100. ARM-based chips sell for about $10, and some chips even cost less than $1. That's why more than 95% of smartphones worldwide use ARM chips.
In other words, on smartphones, Intel competes for rivals with fewer resources than it does. Grove has invited Christensen to go to Intel headquarters in Santa Clara, Calif., to explain his theory of subversion. In retrospect, it seems like this. It was a great irony. Grove later credited this meeting to the main reason why Intel decided to launch the Celeron chip in 1998. Celeron is a cheap product for low-end personal computers, which took up 35 in one year after launch. % market share.
A new wave of gold rush The biggest question now is whether Intel is repeating the mistakes it made on the iPhone in driverless cars. In March last year, it acquired Israel’s digital vision technology company Mobileye for $15.3 billion. In the industry, Intel made a big note this time: With the development of autonomous driving technology, vehicles become computers on wheels, and more and more microchips will be needed. Intel hopes to dominate this field.
Everything Intel has done over the past 50 years has been for high-end, general-purpose chipsets. The company's design and manufacturing processors use an integrated model, which means a lot of fixed costs in research, design, and manufacturing.
The only way to make up for these costs is to sell a lot of equipment with high profits. As a result, Intel is obsessed with technological advances, but is limited by a strict business model. Intel seems to have a monster with a big appetite inside.
However, if autopilot technology does not actually require the computing performance expected by Intel, what is it? This is precisely Huawei's guess. A Huawei executive has revealed that most of Shenzhen's infrastructure will be digitized, and Huawei will pass The 5G network saturates it, which in turn solves the problem of computer speed and latency.
This means that most of the car's computing performance can actually be transferred to the city's infrastructure. This is a radical outlook, but it is clearly a viable option. This means that BMW or Toyota does not need so many high-end chipsets.
Christensen believes that a successful company ultimately fails, not because it is too arrogant, and does not want to change. Kodak, Polaroid and other companies that have finally fallen are deeply aware of the importance of change.
But for these companies, the business model and the needs of existing shareholders have created a very difficult relationship, even the most courageous executives can not cope. Grove once said, 'only paranoia can survive' Maybe he is right.