news

한어Русский языкEnglishFrançaisIndonesianSanskrit日本語DeutschPortuguêsΕλληνικάespañolItalianoSuomalainenLatina

intel

core tips:

1. it took intel only three years to go from being a chip overlord to being an acquisition target.

2. kissinger made a big bet on expensive manufacturing to catch up with tsmc, hoping to restore its former glory.

3. "the market's shift toward ai became the last straw that broke the camel's back for intel."

4. intel is pinning its hopes on the next generation of chip manufacturing technology, but it may be too late.

three years ago, intel was at its peak, with a market value more than twice what it is now, and its ceo pat gelsinger was looking for acquisition targets.

but now, intel itself has become a takeover target. its rapid decline is lamentable. strategic mistakes and the wave of artificial intelligence (ai) have reshaped the fate of this legendary american semiconductor company.

several us media outlets reported that qualcomm, another us chip giant, is making tentative contacts to acquire intel, indicating that intel is experiencing an extremely rare moment of vulnerability in its 56-year history. its problems began with the setbacks the company suffered in chip manufacturing before kissinger took office. as kissinger implemented a costly recovery strategy, the problems worsened.he failed to foresee that the explosive growth of ai would fundamentally shift market demand away from cpus and toward processors developed by rival nvidia.GPU。

"the market shift toward ai over the last two to three years was the final straw for intel. they just didn't have the right capabilities to build ai chips," said angelo zino, a senior industry analyst at independent research firm cfra research.

the former overlord

even if intel is willing to be acquired, a deal with qualcomm is far from certain due to regulatory and other reasons. however, the idea of ​​qualcomm acquiring intel was almost unthinkable not long ago.

for decades, intel has been the world's most valuable semiconductor company, and its chips are almost ubiquitous in pcs and servers. intel is a rare company that integrates chip design and manufacturing, and is a global leader in both.

when kissinger became intel's ceo in early 2021, the company had lost some of its luster and had fallen behind its asian rivals in producing cutting-edge chips.

strategic mistakes

gelsinger, who worked at intel for decades and was the company's first chief technology officer, plans to restore intel to the glory days of its predecessors, including andy grove and paul otellini.

kissinger

to achieve this goal, intel needs to catch up with asian rivals tsmc and samsung electronics in chip production. gelsinger also plans to spend heavily to expand intel's manufacturing business, sell production capacity to companies like qualcomm that only design chips, and enter the foundry business dominated by tsmc and samsung.

it’s an expensive, ambitious gamble, but intel appears to have the ingredients for success: it’s very good at making chips for pcs and servers, and it has a slew of side businesses to fund its next phase of growth.

kissinger quickly set about using intel's financial resources to develop the foundry business, and in the summer after he took office, he started acquisition negotiations with globalfoundries, a global chip foundry, worth about $30 billion. although the deal eventually fell through, kissinger said in an interview in august 2021 that intel still wanted to acquire it. "there will be consolidation in the industry, and this trend will continue. i expect us to be the integrator," he told the wall street journal at the time.

kissinger then finalized a deal to acquire another chip foundry, tower semiconductor, for more than $5 billion. however, the deal was canceled last year because it failed to pass antitrust approval.

gelsinger's goal is to make intel the world's second-largest foundry by 2030, but intel's foundry business got off to a slow start, going through several leaders and many potential customers who limited or canceled business with intel after setbacks in its technology.

the rise of nvidia

as intel's recovery costs mounted, generative ai began to boom. the wave shifted demand from intel's cpus to nvidia's gpus. gpus are designed differently from cpus and are better suited for creating and deploying the most complex ai systems. as tech companies scrambled to get nvidia's scarce ai chips, many of intel's processors sat idle on shelves.

gelsinger was forced to cut costs to sustain his efforts to turn the company around. intel laid off thousands of employees starting in 2022 and cut its dividend last year. it wasn’t enough.last month, kissinger announced that the company would lay off 15,000 employees, cut costs by $10 billion next year, and cancel dividends.

“ai has come much harder than i expected,” kissinger said at the time, calling the layoffs “the hardest thing i’ve ever done in my career.”

nvidia ceo jen-hsun huang

last week, intel announced new initiatives including tighter spending controls and further separation of design and manufacturing, but gelsinger stopped short of selling or spinning off the manufacturing business as some investors have suggested.

"we need to fight for every inch of land and execute more vigorously than ever before,"kissinger told his staff, “because only in this way can we silence our critics and deliver the answers we know we can deliver.”

analysts say the chances of an improvement in intel’s fortunes are narrowing but still possible, and while a falling stock price makes it more vulnerable to takeover bids and activist investors, cost cutting could help the company weather the storm.

as of thursday's close, intel's stock price was down nearly 70% from the beginning of 2020. in 2020, intel's stock price reached its highest point since the bursting of the internet bubble. during this period when intel's stock price plummeted, nvidia's stock price rose more than 18 times. intel's stock price closed up 3.3% on friday after the wall street journal reported that qualcomm was interested in acquiring intel.

maybe it's too late

stacy rasgon, an analyst at bernstein research, said intel's future depends on the success or failure of its next-generation chipmaking technology, which is expected to go into production next year.intel hopes to surpass its competitors at least technologically, and returning to technological leadership will help improve profit margins and enhance customer confidence.

however, intel has a fundamental problem that has not been resolved: its core chip business (cpu) is not expected to recover quickly even as investment in ai chips (gpu) remains strong.

“we can debate whether this strategy is right or wrong, but the problem is that the core business does not support this path,” rasgon said. “however, it may be too late for them now.”

qualcomm

for qualcomm, acquiring intel could help it enter new areas of the chip industry. qualcomm excels at designing mobile phone chips and is a supplier of devices such as apple's iphone. in recent years, it has been building a portfolio of automotive and internet of things chips.acquiring intel would allow qualcomm to build a strong pc and server chip business.

however, it is not clear whether qualcomm will retain intel's manufacturing business in the transaction. unlike intel, qualcomm's current strategy is to outsource production. manufacturing operations are extremely complex and expensive. last year, intel invested $25.8 billion in capital expenditures, equivalent to about 48% of its revenue. in contrast, qualcomm's capital expenditures in the previous fiscal year totaled $1.5 billion, accounting for just over 4% of sales. (author/xiao yu)

for more first-hand news, please download the phoenix news client and subscribe to phoenix technology. if you want to read in-depth reports, please search "phoenix technology" on wechat.