Not only Huawei, Intel also has a “spare tire plan”. Can this save the chip Empire?

There is a famous book in the management field, which is written by Andy Grove, the founder of Intel. This maxim also fully reflects his

. In the book, he said that as a manager, the most important duty is to always guard against other people’s attacks, and pass this sense of prevention to his staff.

if we let go, the strategic turning point will kill the enterprise. Those enterprises that have experienced a turning point and started to decline are rarely able to revive.

recently, Intel announced in the second quarter of 2020 that the chip with 7Nm process technology, which was originally scheduled to be launched at the end of 2021, was postponed by six months.

Intel CEO Bob Swann said that due to defects in its 7Nm process, it was expected that the new product would not be available until the second half of 2022 or early 2023, which is a year behind its internal target.

affected by the epidemic, the growth of PC and Internet services was stimulated. Intel’s performance in the second quarter was very optimistic, with a total revenue of $19.7 billion, a record. Among them, the revenue of PC centered business was 9.5 billion US dollars, with a year-on-year increase of 7%; and the revenue of data center was 10.2 billion US dollars, with a year-on-year increase of 34%.

Stacy Rasgon, an analyst at Bernstein, an investment research firm, wrote in the research report, “to be frank, Intel’s performance figures don’t matter. In fact, investors may read the fourth line of the first page of the financial report and will not read it any more when they see it. ”

this also directly pushed Intel’s share price to dive. At the opening of local time, Intel fell more than 15% and closed at 50.59, down 16.24%. It also hit a new low in nearly four months, with the company’s market value losing $41.5 billion a day to $214 billion.

as soon as the news came out, netizens constantly ridiculed them in China: “no one knows 14nm better than me”, “you don’t understand the romance of 14nm”, “I thought it was squeezing toothpaste before, but later I knew it was really not good”

in 2014, Intel released the first generation of 14nm processor. It was only in 2019 that they launched a 10 nm processor, which took five years. Intel encountered a lot of difficulties in the development of the 10nm process, which was delayed several times, and the first mass production was realized in early 2019.

However, due to the limited capacity of 10nm, the main force of Intel is still 14nm. At present, Intel 10nm process chip is only used for light and thin notebook products with low performance requirements, while high-performance game book and server CPU still use 14nm process chip.

it seems that Intel does not have a good solution for “multiple ticket skipping”. It only said that it will keep a close eye on its 7Nm process development and deliver 7Nm chips on time to make up for the time loss caused by 10nm chip delay.

just over three months ago, at the China annual strategy sharing meeting in 2020, Intel also announced that its chip process improvement speed had been fully restored. Now, the number one player in chip design, manufacturing, packaging and testing, its technical advantages may be slowly disappearing.

on July 28, Intel suddenly announced changes to its technical organization and execution team. Venkata renduchintala, Intel’s chief engineering officer, will leave on August 3, and his technology, system architecture and customer departments will be split into five teams.

according to Intel’s 2020 share entrustment statement, renduchintala is one of the highest paid executives in Intel, with a total annual salary of about $26.88 million as of December 28, 2019.

from design, engineering to manufacturing, he is the head of almost all Intel hardware, and was once considered the most powerful candidate for Intel CEO. It is certain that this internal reorganization is related to the 7Nm delay.

in March this year, Naveen Rao, general manager of Intel’s artificial intelligence business unit, resigned; in May, Craig Barratt, head of Intel’s communication connectivity business department, announced his resignation; in June, Jim Keller, the Silicon Valley star architect who is mainly responsible for the R & D and integration of Intel’s system chips, resigned.

according to foreign media reports, Hagens Berman’s law firm has appealed to Intel investors who have suffered heavy losses to seek possible class action actions, and is also looking for people who can help them investigate potential securities fraud.

in the view of lawyers, since the annual investor conference in 2019, Intel has repeatedly said that it will start shipping the first batch of 7Nm chips in 2021. Because the company claimed that the 7Nm chip will bring twice the area efficiency as the 10nm chip, but the actual situation refused to disclose to the outside world. After the final announcement of the delay, the stock price fell sharply, and it is unlikely to recover in the short term, so there is serious fraud.

it is understood that Hagens Berman law firm is now investigating whether Intel misrepresented and concealed manufacturing and performance problems with its 7Nm chip. They also asked whistleblowers with Intel’s non-public information to contact him. “Under the new scheme, whistleblowers who provide the original information may be awarded a total of no more than 30% of the successful recovery by the sec.”

in 1999, Intel’s market value stood at the highest of US $500 billion, creating the myth of Silicon Valley. In fact, after entering the mobile era, the chip giant who discovered Moore’s law and created an era mark with “Intel inside” has gradually faded away.

Apple computer produced the first handheld computer Newton in 1992, using arm 610 RISC chip.

and Intel has not developed arm chips. In 2002, Intel and Microsoft cooperated to provide Xscale series chips with arm architecture for its handheld computers and smart phones.

after that, Intel sold the Xscale series to Marvell and continued to develop its own x86 processor.

in 2007, after the first generation of iPhone with arm architecture was born, people paid attention to arm architecture. Unlike Intel, arm neither produces chips nor sells chips to end users. Its core business lies in chip technology licensing.

prior to this, Paul Otellini, the former CEO of Intel, had said that he had rejected the opportunity to build a processor for the first generation iPhone because he could not see the future of smartphones.

Intel also missed a good opportunity for development. For a long time since then, Intel has constantly changed its strategy on the mobile market, seeking a breakthrough and cutting into this big cake.

in order to reverse the declining trend of PC decline, Intel has cooperated with OEM manufacturers to promote netbooks since 2008. In 2010, Apple released MacBook Air and launched superbooks.

with apple creating a new category of tablet computers with iPad, Intel also launched the concept of “super book and tablet in one”. But this kind of fuzzy positioning between notebook and tablet computers, coupled with high pricing, has not achieved much market response.

at the same time, Microsoft, a good ally of Intel in the Wintel alliance, has been slow to update windows system both on PC and mobile terminals, which undoubtedly has a negative impact on Intel.

so Intel decided to build its own system. In the face of the attack of iPhone and Android phones, Intel and Nokia hit it off and developed a new system MeeGo. However, only one Nokia N9 mobile phone was launched, and the slogan “don’t follow” has become a masterpiece of this project.

but Intel is still in the arms of Android camp. Subsequently, in 2013, Intel, together with blue magic, Taiwan power, patriot and other manufacturers, launched a tablet computer based on 64 bit low-power chip Bay trail to attract consumers by supporting Android and windows dual system and low price.

but under the pressure of large screen mobile phones, the living space of tablet computers with ambiguous positioning is becoming more and more limited, and even Apple’s iPad is showing a downward trend.

in view of its own process problems, Intel has also announced its “spare tire plan”, which is likely to be TSMC, the world’s largest wafer foundry.

Intel CEO Bob Swann said at the financial report meeting that if the company could not produce 7Nm chips in time, the company would take a more “pragmatic” approach and consider cooperating with third-party wafer factories. This will enable Intel to no longer confine itself to its own factories, but to take into account the capabilities and costs of third-party factories to gain greater selectivity and flexibility.

it is reported that Intel has reached an agreement with TSMC to book the capacity of 180000 6nm chips for TSMC next year. In view of this, xee is likely to have a lot of experience in OEM.

TSMC’s share price has also risen all the way. As of July 28, its total market value has reached US $431.7 billion, becoming the world’s No.1 semiconductor company, and the world’s tenth largest listed company, almost surpassing the two Intel’s.

according to Taiwan media, TSMC does not regard Intel as a long-term customer, so it will not specially expand its production capacity. In fact, TSMC only regards itself as Intel’s “temporary Savior.”.

at present, TSMC’s 7Nm process capacity is quite tight, and 5nm chips have been mass produced since June, and 6nm is only an upgraded version of 7Nm. As the cooperation between Huawei and TSMC will end in September, the spare capacity will also be “scrambled” by other customers. The availability of Intel orders will not have much impact on TSMC.

mosesmann, a securities analyst, said that Intel’s bottleneck in the chip manufacturing process is not just the delay of 7Nm nodes, but the need to restructure the architecture to achieve turnaround, which will cause Intel’s disadvantage in the manufacturing process for five, six or even seven years.

Chang i-chien, an analyst at Taishin securities investment advisory, said that in the long run, it would be more cost-effective for Intel to outsource chip manufacturing to a foundry. In the future, Intel may gradually reduce its own chip production. If one day Intel sells its U.S. factory, it’s no surprise to the chip industry.

the traditional IDM model was once regarded as a successful example of IT industry. Early born enterprises, almost without exception, choose it, from design to manufacturing, from technology to brand, almost all of them. The advantage of the first mock exam

is that resources can be highly controlled and can be integrated into a seamless whole. It has created the glory of IBM, Motorola, Siemens, HP, Philips and many other companies.

but it is obvious that all of these brilliant enterprises have been deeply trapped in the IDM mode, and some of them have started to walk out of IDM and choose open cooperation after encountering serious crisis.

today, with the development of the semiconductor industry chain, the division of labor in the industry chain is more refined and differentiated. It does not mean that a company is doing the whole thing from the beginning to the end. Therefore, it will be more wise for Intel to focus on chip design with its own advantages.

in fact, amd also used the traditional IDM mode. In order to better compete with Intel, amd made a strategic change in 2008, choosing to peel off chip manufacturing and move towards professional design mode.

this also proves that IDM is not a once and for all business

Author: zmhuaxia