Stock price hits 10-year high, redesigning 7nm, key point of Intel's transformation

Under the dual heat of the strong return of 30-year veteran Pat Gelsinger and the 2020 financial report that exceeded expectations, Intel's stock price soared all the way at the end of today's trading, and finally closed at US$62.46. The market value reached US$255.961 billion, close to the historical high in the past 10 years.

In a conference call after the earnings report, Intel's current CEO Bob Swan made a rare frank admission that "there were some problems when Intel's 7nm technology was initially defined."

Bob also said that after announcing the delay of 7nm technology in July, Intel began re-architecting these steps and streamlined the process architecture of 7nm chips to ensure that the product can be successfully launched in 2023.

At a time when its process technology is frequently frustrated and its data center business is surrounded by competitors, the return of veteran Pat Gelsinger undoubtedly indicates that Intel has now reached a critical juncture in its transformation.

1. Revenue and profit declined for three consecutive quarters

The financial report shows that in 2020, Intel's annual revenue was US$77.9 billion, an increase of 8% year-on-year; Q4 revenue was US$19.978 billion, a decrease of only 1% year-on-year.

Among them, the PC business grew by as much as 33%, and the PC-centric CCG department's annual revenue reached US$40.1 billion, an increase of 8% year-on-year.

All Intel shareholders, big and small, finally breathed a sigh of relief.

Intel's quarterly revenue and profit in 2020

After facing a decline in both revenue and profits in Q1, Q2 and Q3, and management's dismal Q4 revenue forecast (a 14% drop), this decent financial report finally reassured the market a little - although it is hard to say whether this was a "financial cleansing" prepared by Intel for the new CEO before he takes the stage in February.

In the conference call after the financial report, Intel's current CEO Bob Swan made a rare further positive response to the problems and progress encountered in the 7nm process.

In July this year, Intel announced that the 7nm CPU originally scheduled to be launched at the end of 2021 will be postponed to 2023 at the latest; in contrast, the 7nm products of its "old rival" AMD have already been on the market for several months.

2. The PC market has “branched”

During the COVID-19 pandemic that swept across 2020, the demand for remote work, online education, personal entertainment, etc. surged, and the already highly saturated personal computer (PC) market finally regained its momentum after years of lukewarmness.

Data of the world's top five PC manufacturers in 2020

In 2020, Lenovo, HP, Dell, Apple, and Acer, the world's five largest PC giants, all reported high year-on-year growth - according to IDC data, global PC market shipments in 2020 increased by 13.1% year-on-year, and Q4 growth reached an astonishing 26.1%.

This is mixed news for Intel.

The good news is that if there had not been this epidemic, as the PC industry has been hovering at the high point of market saturation for many years, according to Intel's expectations when the 2019 financial report was released, the PC-based Client Computing Center (CCG) business revenue would not only have "almost no profit", but the growth in 2020 would also drop below single digits.

In fact, Intel's CCG division generated $40.1 billion in revenue this year, up 8% from $37.1 billion last year.

However, this also means that Intel's CCG business still accounts for more than 51% of its full-year revenue, the same as last year.

Intel's pan-data center business revenue share from 2016 to 2020

However, as early as 2015, Intel began its data-centric transformation.

3. Data center market surrounded by strong competitors

Currently, Intel's two main business camps are divided into PC business (Client Computing Group CCG) and pan-data center business (Data Center Group DCG, Internet of Things Department IOTG, Mobileye, Memory Group NSG, and Programmable Solutions Group PSG)

In order to form the current big family of the Pan-Data Center Business Unit, Intel has continued its "buy, buy, buy" business for many years.

For example, the establishment of Intel's Programmable Solutions Group (PSG) can be attributed to its acquisition of FPGA giant Altera for US$16.7 billion in 2015.

Mobileye, which was acquired for US$15.3 billion in 2017, contributed more than 10% of Intel's full-year year-on-year revenue growth in Intel's 2020 financial report, with Q4 growth as high as 39%.

And in the $3.8 billion in artificial intelligence business revenue announced by Intel, how can we miss the neuromorphic chip company Nervana, AI terminal vision chip company Movidius, and AI cloud training chip company Habana, which it acquired in 2016 and 2019 respectively?

It's good to have money.

Although it missed the big wave of the mobile phone era, Intel's X86 architecture's overwhelming dominance in data center CPUs still allowed it to enjoy high market recognition when it first launched its data center transformation plan.

In 2016, Intel's pan-data center business grew by 34.64%, and the market was optimistic about it.

However, as other players such as AMD and Nvidia have gradually set their sights on this trillion-dollar piece of fat, Intel's data center business growth has slowed rapidly, to 15.77% in 2017, 20.43% in 2018, and even fell to 2.91% in 2019, before rebounding to 8.06% in 2020.

The picture comes from the Internet

Needless to say, NVIDIA, the "new wave" that rose on the AI ​​craze, is challenging the development of Moore's Law with "Huang's Law" day after day.

With the natural advantages of GPU technology and NVIDIA's keen early investment in data centers, high-performance computing, artificial intelligence and other fields, NVIDIA's stock price has doubled since 2016, and its share price has continued to rise with the artificial intelligence craze.

Nvidia stock price

While the industry was still using NVIDIA V100 as the comparison standard for AI chips, at last year's GTC, NVIDIA launched the A100 GPU, which had 20 times the performance of V100, achieving the largest performance leap in NVIDIA's history.

Although 36Kr learned from industry sources that Nvidia's main sales force in data center products is still V100, the launch of A100 is still "not very harmful but has a strong deterrent effect" on most AI chip companies.

On July 9 last year, an ordinary Wednesday, Nvidia's stock price rose 3.49% to $408.64, with a total market value of approximately $251.314 billion, surpassing Intel for the first time to become the world's most valuable chip company.

——Also in July last year, Intel officially announced the postponement of 7nm.

Nvidia didn't stop here.

Since July last year, Nvidia has once again released several blockbuster news, including its $40 billion acquisition of Arm, which shocked the global industry. Since then, Nvidia's stock price has soared again, and has remained above $500 since August this year, with a market value of more than $340 billion, far surpassing Intel.

However, its "old enemy" AMD, under the leadership of Chinese-American female CEO Lisa Su, made a strong comeback in six years. Not only did it successfully pull AMD back from despair, but it also achieved a crazy counterattack in market value and stock price. The stock price soared from the lowest of less than $2 six years ago to more than $90 today, with a market value of over 100 billion.

Stock Price Data

Unlike Intel's widely criticized "toothpaste squeezing", AMD has made rapid progress in the field of advanced processes. Not only did it launch the world's first 7nm GPU in 2018, it has also been making rapid progress in the Ryzen, EPYC and other processor markets, and has achieved year-on-year revenue growth of more than 25% for several consecutive quarters.

In just a few years, AMD has dramatically increased its market share in the data center from no less than 1% to more than 10%, and has successively won large orders from AWS and Oracle.

The picture comes from the Internet

Similar to Nvidia's deal with Arm, last year AMD also announced that it would acquire Xilinx, the world's largest FPGA chip leader, for US$35 billion, and the merger is expected to be completed by the end of 2021.

After acquiring Xilinx, AMD will not only gain FPGA market coverage similar to Intel's acquisition of Altera, but will also gain technical support from Xilinx's latest ACAP architecture and Deephi Technology, a Chinese AI chip startup previously acquired by Xilinx, in the areas of AI chips and in-vehicle intelligence.

After the merger, AMD will have 13,000 engineers and annual R&D investment of more than $2.7 billion.

Before Intel's CEO change, facing market competition surrounded by strong rivals, Intel was frequently controversial. Last year, when global semiconductors were generally booming, its stock price fell by nearly 13%.

4. Pat Gelsinger

If Intel had a nickname system, Pat Gelsinger would be called "Instant Rain".

After the official announcement that Intel's 30-year veteran would replace current CEO Bob Swan and officially return to lead Intel, the market gave it face and soared 11% at the opening, sweeping away the gloom that had lasted for several months.

Rather than saying that the market misses Pat Gelsinger, it is better to say that the market misses what Pat Gelsinger represents, the era when Intel was dominating the world and leading the technological revolution.

During today's earnings call, Pat Gelsinger, who will take over in February, set the tone for his return with a simple opening statement.

“I joined Intel when I was 18 years old, and I’m proud to say that over the next 30 years, I’ve learned a lot from industry giants like Andy Grove, Gordon Moore, and Robert Noyce.”

Among the names cited by Pat Gelsinger, Gordon Moore is the founder of Moore's Law and Robert Noyce is the father of integrated circuits.

——Andy Grove is the legendary CEO who boldly led Intel to completely transform into the computer processor market and achieved Intel's highest glory.