news

한어Русский языкEnglishFrançaisIndonesianSanskrit日本語DeutschPortuguêsΕλληνικάespañolItalianoSuomalainenLatina

during the midday trading of the u.s. stock market on friday, september 20, eastern time, the wall street journal quoted an insider as saying that qualcomm had recently made an acquisition offer to intel and had contacted intel about the acquisition. subsequently, cnc also confirmed the news. this is the second time this month that the media has reported that qualcomm is interested in acquiring intel.

influenced by the news, intel's stock price, which had fallen by more than 3.7% in the midday trading of the us stock market on friday, rose sharply. when the stock price reached a high of $23.12, the intraday increase was nearly 9.4%, and the highest increase was about 13.6% from the intraday low of $20.35. due to the large fluctuation of the stock price, it also triggered a trading suspension. although it fell back in the late trading, intel still closed at $21.84, an increase of 3.31%, and its market value rose to $93.388 billion.

in contrast, qualcomm's stock price fell rapidly during the trading session due to the news. in just about 10 minutes, the stock price plummeted to us$164.30, and the decline widened from less than 2% to 5.5%. in the late trading, the decline narrowed to 2.87%, closing at us$168.92, with a market value of us$188.2 billion.

the wall street journal also pointed out in its report that people familiar with the matter warned that the acquisition deal is far from certain. even if intel is willing to accept it, such a large-scale transaction will almost certainly be subject to antitrust review. however, the deal may also be seen as an opportunity to enhance the united states' competitive advantage in the chip field. in order to reach a deal, qualcomm may intend to sell some of intel's assets or parts of its business to other buyers.

regarding the rumors of qualcomm's acquisition of intel, even if the news is true, core intelligence believes that the management, which is actively promoting intel's reform, will most likely not agree to the deal, at least not without a high enough premium. personally, i think it may take more than $200 billion for them to be tempted, but qualcomm may not be able to afford this price. even if the deal is reached, in the context of the united states' revitalization of its domestic semiconductor industry, the us antitrust department may be inclined to approve the deal, but the eu and chinese regulatory authorities will most likely not approve it. and if the deal fails, qualcomm may have to pay a large breakup fee. therefore, this acquisition is indeed a bit unreliable!

intel management's "ambition"

although intel released poor second-quarter earnings and forecasts in early august this year, and announced a 15% global layoff, cut capital expenditures (us$10 billion in capital expenditures by 2025), and suspended quarterly dividends (the first suspension of dividends since 1992), intel's stock price plummeted by more than 21% in a single day, and intel has also been under tremendous pressure from investors, the government, internal employees, social opinion, and many other aspects. it can be said that intel is in the midst of a major crisis since its founding 56 years ago.

this has also triggered a series of rumors such as "intel will sell its wafer manufacturing business", "intel will sell its fpga (altera) business", "intel will suspend overseas wafer factory construction", and "qualcomm will acquire intel".

however, after the subsequent intel shareholders' meeting, intel ceo pat gelsinger confirmed in an open letter that the construction plans of the german and polish factories would be postponed, but he also denied the rumors of selling the wafer manufacturing business and completely selling the fpga business. he restated intel's accelerated transformation plan and announced the good news of obtaining cooperation with amazon aws and a $3 billion subsidy from the us department of defense.

as the helmsman of intel, kissinger said bluntly: "this is the most important transformation for intel in more than 40 years. we have not attempted anything so important since the transition from memory to microprocessors. we succeeded then, and we will meet this moment and build a stronger intel in the decades to come."

as early as the beginning of 2021, when intel invited back veteran kissinger to take the helm again, the intel board of directors had high hopes for kissinger, hoping that such a technical veteran could lead intel to glory again.

after kissinger officially took office as ceo of intel, he carried out drastic reforms on intel and proposed the "idm 2.0" strategy, focusing on core advanced process manufacturing capabilities to enhance product strength, while unbundling the design business from the manufacturing business (opening up outsourcing), expanding production capacity, and developing wafer foundry business.

obviously, this "idm 2.0" strategy is not a strategy that can be effective in the short term. after all, intel and tsmc were in an embarrassing situation of lagging behind in process technology at the time. at the same time, the global expansion of wafer fabs itself has a long cycle, and the transformation of an idm manufacturer's manufacturing business department into a wafer foundry that can compete with tsmc is even more impossible to achieve overnight.

as of now, kissinger has been in charge of intel for more than three years. his proposed "four advanced process nodes in five years" plan is about to be completed as scheduled. although intel missed the huge opportunities brought by the "ai data center" market in the process, its share in the core data center market is also being squeezed by competitors such as nvidia, amd, and arm.

according to data from the second quarter of 2024 released by market research firm mercury research, amd's market share in the three major cpu categories of servers, desktops and mobile has increased by several percentage points compared with the same period last year, while intel has declined. however, even so, intel still occupies about 70% of the x86 cpu market, which currently accounts for 90% of the data center market share.

in addition, intel has been returning to the gpu track since 2020, and its gpu business has been growing steadily. with its leading position in the data center cpu market, its data center gpu business has broad prospects. intel also plans to turn the next generation of ai accelerators to gpu architecture in order to better compete with nvidia.

for kissinger, the intel 18a process, which intel will mass produce in the first half of 2025, will be the key to turning things around. not only will intel's next-generation pc and data center products return to internal manufacturing and adopt the latest intel 18a process, but intel also hopes to win more foundry customers with intel 18a.

recently, in order to ensure the smooth mass production of intel 8a process in 2025, intel also announced that it would "skip the productization" of intel 20a node and invest engineering resources from intel 20a to intel 18a in advance. the process will adopt the ribbonfet all-around gate transistor architecture and powervia backside power supply technology that have been completed on intel 20a. according to the data released by intel, the defect density of intel 18a has reached the d0 level, which is less than 0.40 (def/cm^2). this indicator means that the intel 18a process node is generally considered to be production-worthy and in good operating condition.

intel has announced that microsoft plans to use intel's 18a process node to produce a chip designed by it. intel will also use the intel 18a process to produce ai fabric chips for aws.

it is foreseeable that if intel's 18a process is successful, it will not only help intel's own products expand their competitiveness, but will also successfully open up the situation for intel's wafer foundry business. according to intel's forecast, it will generate some wafer foundry revenue from external customers in 2026 and generate "meaningful" revenue in 2027.

in addition, suspending the construction of the german wafer factory and the polish packaging factory, which cost nearly 35 billion euros in total, for two years will greatly ease intel's capital expenditure pressure in the next two years. at the same time, intel's factory construction plan in the united states continues to advance, and it is expected to obtain the $8.5 billion direct subsidy funds and $11 billion loan that have been agreed with the us department of commerce. coupled with the latest announcement of $3 billion in funding from the us department of defense, these funds will help intel significantly ease the funding needs of current and subsequent factory projects under construction.

it can be said that the crisis intel is currently facing is the "pain" it must face in the process of transformation, but this crisis has not yet reached the point of "life and death". various external public opinions and hype may have exaggerated the "sense of crisis" too much.

intel still has a lot of cards to play. whether it continues to promote altera's ipo or completely sells altera, it can recover a large amount of funds and solve the current crisis. after all, when intel acquired altera in 2015, it spent a huge amount of us$16.7 billion. in 2022, amd spent more than us$35 billion to complete the acquisition of xilinx (later, as amd's stock price rose, the final transaction value greatly exceeded the initial price of us$35 billion).

in addition, intel can also follow amd's example of divesting its wafer manufacturing business (which later became globalfoundries) and completely sell its wafer manufacturing business to solve the current crisis. however, since advanced process manufacturing capabilities are the key to the idm 2.0 strategy, this possibility is not great. however, when the wafer manufacturing business is independently operated and on track, intel will most likely promote an independent ipo of the business to recover funds and reduce risks.

from the current perspective, if the board of directors continues to trust kissinger, then kissinger will continue to promote the completion of intel's strategic transformation during his tenure at the helm of intel to fulfill his promise to "build a stronger intel in the coming decades." of course, if the transformation is still ineffective, kissinger may face "dismissal" at that time. but before that, as a technical veteran who has worked at intel for more than 30 years, kissinger will never agree to sell the company to other competitors at a low price during intel's major transformation process at a historical low stock price since 2015.

what is the attitude of the board of directors?

looking back at the tenures of several intel ceos over the past two decades, the three ceos before kissinger - paul o'lorning (2005-2013), brian krzanich (2013-2018), and bob swan (2018-2021) - did not serve for long periods of time, and their tenures have been shortened. this also reflects the increasingly fierce market competition. however, it should be pointed out that two of the three ceos studied business administration, and only krzanich had some technical background.

the reason why intel's chairman's board brought back henry kissinger, a veteran intel executive with a technical background, to take the helm is naturally because they have high hopes for him and hope that such a senior technical ceo can bring intel to greater glory. however, if he is not given enough trust and the time period to execute his transformation plan, it may be difficult to reverse intel's decline by changing another ceo, and in the end, intel may only be able to "sell itself."

from the perspective of maximizing profits, intel's board of directors is unlikely to be willing to sell the company at a low price when intel's stock price is at its lowest point in nearly nine years and its transformation is about to be completed.

let's take a look at intel's current board members and their backgrounds:

1. tsu jae king liu, dean of the college of engineering and roy w. carlson professor of engineering at the university of california, berkeley (2018-present). distinguished professor of electrical engineering and computer science at tsmc (2014-2018); co-founder and president of progressant technologies, a startup developing negative differential resistance transistor technology (2000-2004).

2. risa lavizzo mourey, university of pennsylvania, robert wood johnson foundation pik professor, population health and health equity (2018-2021 (retired)). president and ceo (2003-2017) of the robert wood johnson foundation, the largest healthcare-focused philanthropic organization in the united states. sylvan eisman professor of medicine and health systems at the university of pennsylvania school of medicine (1995-2001).

3. james j. goetz, holds a bachelor's degree in electrical engineering from the university of cincinnati and a master's degree in electrical engineering from stanford university. he is a partner at sequoia capital operations llc (2004-present); co-founder of software design, development and strategy company vitalsigns software (1996-1999); director of data security and storage company barracuda networks (2009-2017), wireless (wi-fi) network equipment manufacturer ruckus wireless inc. (2012-2015); and director of network security solutions company palo alto networks inc.

4. alyssa henry, holds a bachelor's degree in applied science (computer science) from the university of california, los angeles. she was the ceo of block co., ltd. (formerly square), which provides software, hardware, and financial services to small businesses and individuals (february 2023-october 2023 (retired)), and the vice president of amazon web services storage services at amazon (2006-2014). prior to amazon, she worked at microsoft corporation for 12 years and has extensive experience in engineering, project management, and product unit management.

5. dion j. weisler holds a bachelor of applied science in computing from monash university, australia. he served as president and ceo of hp inc. (november 2015 to october 2019). prior to joining hp in 2012, he served as vice president and chief operating officer of products and vice president and general manager of the mobile internet digital home group at lenovo group (2007 to 2011). his previous experience also includes management positions at telstra corporation limited (a telecommunications company) and acer incorporated (a hardware and electronics company).

6. stacy j. smith has served as senior vice president and chief financial officer of intel corporation since 2010. he joined intel corporation in 1988 and served as vice president of sales and marketing in 2002. he was appointed vice president, chief financial officer, chief business services and chief information officer in may 2004. he was appointed vice president and assistant chief financial officer in march 2006 and vice president and chief financial officer in october 2007. he served as a director of jiwo corporation in june 2010.

7. frank d. yeary has served as a director of intel since 2009. since 2012, he has served as ceo of darwin capital advisors, llc, a private investment and consulting firm in san francisco, california. from 2008 to 2012, he served as vice president of the university of california, berkeley, and from 2010 to 2011 he served as interim chief administrative officer, managing financial and operating responsibilities. before returning to his alma mater in 2008, he spent nearly 25 years in the financial industry, most recently at citigroup investment bank, financial services company as a managing director, global head of mergers and acquisitions and a member of the management committee. he is also chairman of the consulting firm camberview partners, llc.

8. omar ishrak has served as chairman and ceo of medtronic since 2011. prior to joining medtronic, he served as president and ceo of ge healthcare systems, a comprehensive provider of medical imaging and diagnostic technologies, from 2009 to 2011. prior to that, he served as president and ceo of ge healthcare clinical systems from 2005 to 2008, and president and ceo of ge healthcare ultrasound and bmd from 1995 to 2004.

9. gregory d. smith, formerly served as boeing’s chief financial officer and executive vice president of corporate operations (2020-2021 (retired)), chief executive officer and chief financial officer (december 2019 to january 2020), chief financial officer and executive vice president of corporate development and strategy, and executive vice president and chief financial officer (2012-2015). vice president of global investor relations at raytheon (2004-2008).

10. intel ceo henry kissinger is also a director of intel’s board of directors.

judging from the resumes of the above-mentioned board members, most of them are from engineering backgrounds and have management backgrounds in well-known electronic technology companies. at least for now, most of the board members still trust kissinger and approve of his transformation plan. even though intel is currently in trouble and has aroused external associations, various rumors about intel are flying around, but there has been no rumor that the intel board of directors is considering impeaching kissinger.

it is worth noting that on august 22, intel announced that lip bu tan (chen lip-wu), a veteran in the semiconductor industry, withdrew from the intel board of directors for personal reasons. later, the media broke the news that lip bu tan withdrew because of disagreements with the ceo and board executives on many issues (mainly about mid-level layoffs). this also reflects that most members of intel's board of directors still support kissinger's transformation strategy for intel.

a familiar hunting scene

as early as november 2017, chip design giant broadcom made an acquisition offer to qualcomm for $70 per share, totaling approximately $105 billion. the specific payment method was $60 in cash plus $10 in broadcom equivalent stock. the acquisition offer at that time also agreed that qualcomm would continue to acquire nxp semiconductors for $38 billion and was willing to assume $25 billion in debt after the acquisition. the premium was more than 20% relative to qualcomm's market value at the time.

qualcomm at that time, like intel now, was in deep trouble.

at the end of 2016, qualcomm was first fined 1 trillion won (about 5.9 billion yuan) by the korea fair trade commission for monopoly and obstruction of competition, setting a record for antitrust fines in south korea. in early january 2017, the us federal trade commission also filed a lawsuit against qualcomm, accusing qualcomm's patent licensing policy of violating federal laws and constituting unfair competition. subsequently, the conflict between apple and qualcomm over patent licensing fees broke out, and apple stopped paying patent licensing fees to qualcomm in april 2017. subsequently, qualcomm and apple broke out in a protracted patent lawsuit around the world. at the same time, qualcomm's $38 billion acquisition of nxp semiconductors has not yet fully obtained the approval of china's antitrust department (it was only announced to be abandoned in july 2018). this series of problems caused qualcomm's profits to plummet and its stock price to continue to fall.

according to qualcomm's fourth quarter financial report for fiscal year 2017 released in early november 2017, due to apple and its suppliers not paying patent licensing fees and other legal settlements reached by qualcomm, qualcomm's revenue fell 5% to us$5.9 billion, while profits plummeted 89% to us$168 million.

just when qualcomm was stuck in the "quagmire", broadcom joined forces with some consortiums to start hunting qualcomm. however, even in february 2018, when broadcom raised the acquisition price per share to $82, a 17% increase from the previous $70, and the total acquisition offer to $121 billion, and launched a "proxy battle" (arranging new directors through shareholders' meetings to control board seats), qualcomm's management still firmly opposed it.

at the same time, in order to fight back against broadcom's acquisition, qualcomm management not only rejected broadcom's director nomination, but also set up new obstacles for broadcom's hostile takeover. once the company's control changes, if existing employees are fired, the acquirer must pay higher severance pay. at the same time, qualcomm also spoke out with major partners, and many partners said that they "are not optimistic about broadcom's acquisition of qualcomm, because this will bring uncertainty to these partners' future cooperation with qualcomm."

faced with broadcom's relentless pressure, according to media reports at the time, qualcomm told broadcom in february 2018 that it would agree to the deal if the other party's bid was as high as $160 billion (including debt). compared with the initial acquisition offer of about $105 billion, the premium was more than 50%. at the time when broadcom made the acquisition offer, its market value was only about $100 billion. this also made it difficult for the two sides to reach an agreement.

finally, on march 12, 2018, us president donald trump signed an executive order banning broadcom from acquiring qualcomm on the grounds of threatening us national security, and the farce finally came to an end.

judging from the subsequent results, the decision of qualcomm management was undoubtedly correct. the previous lawsuit between qualcomm and apple was finally settled in 2019 with apple "admitting defeat". apple continued to purchase baseband chips from qualcomm and pay patent fees. as of now, apple is still completely dependent on qualcomm's baseband chips.

in terms of market value, as of september 20 this year, qualcomm's market value has reached us$188.177 billion, far higher than broadcom's initial offer of us$121 billion.

in comparison, the difficulties intel is currently facing may not even be as severe as the "difficulties" qualcomm was facing at the time. intel's stock price has fallen from more than $216 billion at the end of 2023 (the stock price peaked at $50.60) to the current $93.388 billion (closing price on friday was $21.84), a drop of more than 56%. intel's management and major shareholders do not agree with intel's current undervalued market value.

according to the sec documents, on august 5 this year, kissinger purchased 7,500 shares of intel common stock on the open market at an average price of $20.3137 per share through his family trust, and on the same day purchased 5,000 shares of intel common stock on the open market at an average price of $19.9186 per share. although the increase in shares is not much, it also reflects kissinger's belief that the company's stock price is seriously undervalued.

in addition, according to data, as of the end of the first quarter of this year, vanguard group, blackrock, state street corporation, capital international, and geode capital management llc were among intel's top five institutional shareholders, and these five shareholders have been increasing their holdings of intel shares since the first quarter of 2023.

obviously, if qualcomm cannot make a sufficiently attractive acquisition offer, intel's major shareholders and management will not agree to the acquisition. chip intelligence predicts that if the acquisition offer is less than $200 billion, intel may not even consider it. however, qualcomm's current market value is only $188.177 billion, which makes the so-called acquisition difficult to achieve.

if the rumors that qualcomm plans to acquire intel are true, then the subsequent game between qualcomm and intel management will probably be no less than the game between broadcom and qualcomm management.

changing attitudes of regulators

judging from the revenue scale and market value of qualcomm and intel, if the two companies can really merge, they will create a chip company with a market value of 300-400 billion us dollars, with annual revenue exceeding 100 billion us dollars, becoming the world's largest semiconductor company.

if the two companies complete the merger, their businesses will become an all-around giant in the fields of data centers, consumer electronics (including desktop and mobile), automotive electronics, industrial control, the internet of things, and wired and wireless networks. at the same time, the two parties will form complementary synergies in many markets, further increase their market share, and even form a monopoly advantage.

for example, in the pc market, the merger of intel and qualcomm will become the leader in both x86 pc and arm pc. if qualcomm develops arm server chips in the future, then with the help of intel's current dominant position in the server cpu market, qualcomm may become the leader in the arm server cpu market. although there will be competition between the two businesses, there are already two different demands in the market. it is better to do it yourself than to let others grab it. in the automotive electronics market, qualcomm currently monopolizes the global high-end smart cockpit market, while intel's mobileye has a strong advantage in the adas market. the merger of the two parties will help further expand its share in the smart cockpit and autonomous driving fields. in addition, qualcomm's technological advantages in communication baseband chips and wireless fields can also bring more empowerment to intel in more markets. more importantly, qualcomm, as the world's leading chip design manufacturer, originally mainly outsourced its chips to tsmc and samsung. if it merges with intel, it may completely turn to intel's internal foundry in the future, which will undoubtedly promote the rise of intel's foundry business. as a result, it is even possible to surpass tsmc.

of course, the merger of intel and qualcomm is something that other competitors do not want to see. at the same time, it will also touch the sensitive nerves of regulators in various countries.

unlike the us government's ban on broadcom's acquisition of qualcomm on the grounds of "threat to national security", this time it is a merger of two purely american companies, and it is against the backdrop of the united states vigorously developing its domestic semiconductor industry. if qualcomm and intel reach an acquisition deal, us regulators may approve the transaction.

however, antitrust agencies in the european union, china and other places are unlikely to agree to this acquisition.

looking back at several major mergers and acquisitions in the semiconductor field in recent years:

qualcomm's $38 billion acquisition of nxp semiconductors was not approved by china's state administration for market regulation before the deadline;

sk hynix's acquisition of intel's flash memory business was conditionally approved by china's state administration for market regulation;

amd's $35 billion acquisition of xilinx was conditionally approved by china's state administration for market regulation;

intel's $5.4 billion acquisition of tower semiconductor was rejected by china's state administration for market regulation.

obviously, as competition between china and the united states in the semiconductor field becomes increasingly fierce, chinese regulators have begun to tighten their review of large-scale mergers and acquisitions of foreign semiconductor giants and have begun to use their veto power.

therefore, even if qualcomm really reaches an acquisition deal with intel and obtains approval from the united states and the european union, china will not allow these two american semiconductor giants to strengthen their monopoly advantages in many global markets through mergers unless the us government offers sufficient benefits in exchange.

author: core intelligence - rurouni sword