Intel is going to lay off employees again.
In early August, Intel’s performance was announced at the same time as a notice of 15,000 layoffs.
Intel CEO Pat Gelsinger said this was part of the company’s $10 billion cost-cutting plan. “I have no illusions that the road ahead of us will be smooth. Neither should you. Today is a hard day for all of us, and there will be more hard days in the future.”
But Intel did not give a specific detailed layoff plan. Intel told the Times Weekly reporter: “This reduction measure is global.
The company will not announce the number of employees affected by specific regions or locations.”
Intel said that Intel is working hard to accelerate its strategy while significantly reducing costs. It will reduce costs and improve efficiency through a variety of measures, including cutting positions in some businesses and functional departments across the company.
Intel (INTC.O) 2024 second quarter financial report data shows that the second quarter revenue was $12.8 billion, down 1% from the previous year, with a gross profit margin of 38.7%; under GAAP (generally accepted accounting principles), the net profit loss was $1.6 billion.
Intel’s wafer foundry, which has high hopes, had revenue of $4.3 billion, down 1% from the previous month.
The release of this second quarter report brought a series of shocks. First, Intel’s stock price fell by 26% to $21.48 on Friday.
According to media reports, this is the largest single-day drop in Intel’s stock since 1982.
On the same day, most U.S. technology stocks fell. The three major indexes all closed sharply lower, with the Dow Jones Industrial Average closing down 1.21%, the Nasdaq closing down 2.30%, and the S&P 500 closing down 1.37%.
“Despite our progress in products and processes, profitability in Q2 was disappointing,” said Intel CEO Pat Gelsinger at the conference.
On the same day as the earnings report was released, a notice appeared on Intel’s official website. Pat Gelsinger said that Intel will lay off at least 15% of its employees as part of the company’s $10 billion cost-cutting plan.
Pat Gelsinger believes that “not fully benefiting from powerful trends such as artificial intelligence” is one of the reasons why Intel’s revenue is lower than expected.
In order to catch up with its competitors’ first-mover advantage in this round of AI wave, Intel has put a lot of effort into AI in recent years. From betting heavily on AI chips to splitting the wafer foundry business line, the CPU overlord is eager to “become king” again in the AI ​​era.
The blue giant transforms to reduce costs and increase efficiency
Silicon Valley’s “blue factory” is in a dilemma.
Intel’s CEO admitted this. He believes that at a time when technological innovation has repeatedly achieved results, the company is facing the dilemma of “too high costs and too low returns.”
Intel’s prescription for itself is “cost reduction”. On the same day as the financial report was released, Intel proposed a plan to reduce costs by $10 billion by 2025. Specific actions include layoffs, simplifying product portfolios, reducing capital expenditures, suspending dividends, etc.
According to the financial report, as of December 30, 2023, Intel has a total of 124,800 employees worldwide. Data from the same year showed that Intel has more than 12,000 employees in China.
According to the financial report, as of December 30, 2023, Intel has a total of 124,800 employees worldwide. Data from the same year showed that Intel has more than 12,000 employees in China.
According to China Electronics News, the Chinese market brings about 25-30% of Intel’s revenue. Tianyancha data shows that the registered locations of Intel China Holdings’ subsidiaries are Beijing, Shanghai, Dalian, Chengdu, and Hainan.
If the layoff indicators are “fairly distributed” according to the region, the number of layoffs at Intel China will be around 1,500.
Times Weekly reporters found that users who claim to be Intel employees and talk about layoffs on social media come from different departments and have different IP addresses.
Faced with the concerns of layoffs, some users said they had already given up “waiting for the gift package (layoff compensation)”, while others were tired of the “blunt knife cutting meat” style of continuous layoffs.
There are also users who claim to be former employees who post that Intel has come to this point due to the international situation, business decisions and other factors.
However, most users who speak out through social platforms are unwilling to speak out about layoffs. The Times Weekly reporter tried to contact them through the platform, but did not get more replies.
Intel has had multiple rounds of layoffs in the past two years. According to posts by users on social platforms who claim to have experienced layoffs, although they feel complicated about leaving the company, they are relatively satisfied with the “layoff gift package”.
Regarding details such as layoff compensation, business lines involved, and regions involved, the Times Weekly reporter asked Intel for confirmation.
But Intel said it would not announce the number of affected employees according to specific regions or locations. “We will treat every employee with respect and do our best to provide help and support,” Intel told the Times Weekly reporter.
Intel expects that by the end of 2024, the company’s headcount will be reduced by 15,000 or more than 15%, most of which will be completed before 2025.
Lightweight and simple will certainly bring speed, but the cold has already spread to the surrounding areas. “In the short term, it can indeed quickly relieve financial pressure, optimize organizational structure, improve organizational efficiency and decision-making speed, but it may also lead to low morale and partial loss of technology.” Guo Junli pointed out.
Another means for Intel is to simplify the product portfolio. In Guo Junli’s view, simplifying the product portfolio can make the business more focused on products and technologies with strong profitability and good market prospects, while reducing R&D, production, market and other costs, and enhancing overall competitive advantages, but there is also a risk of losing some market segments and losing some customers due to less market coverage.
Intel expects that after a multi-pronged approach, total capital expenditures in 2024 will be reduced by more than 20% from previous forecasts, totaling US$25 billion to US$27 billion; this figure will be US$20 billion to US$23 billion in 2025.
Under the wave of AI, the progress is slightly slow
Where did Intel spend its money, what results did it achieve, and where did the cost go?
Overall, the company’s product department revenue in the second quarter was US$11.8 billion, up 4% year-on-year.
The traditional strong business CCG was driven by the AI ​​PC concept and performed relatively well, with revenue of US$7.4 billion, a year-on-year increase of 9%.
At the performance report conference, Intel revealed that AI PC shipments have exceeded 15 million units, “several times more than the sum of our competitors.” It is expected that more than 40 million units will be shipped by the end of 2024, and cumulative shipments will exceed 100 million units by the end of 2025.
In addition to the CCG business, data center and AI business are also the main sources of the company’s revenue. In the 2023 annual report, the combined revenue of these two business lines accounted for at least 80%.
Under the AI ​​tide, the business demand of data centers has been pushed up, but Intel’s performance is not excellent whether compared with previous performance vertically or horizontally in the industry.
In the second quarter, its DCAI business revenue was US$3 billion, a year-on-year decrease of 3%.
In the short term, the sluggishness of the data center business may have something to do with the fact that Intel’s new products have not yet been officially launched.
At the report conference, Intel announced that the Gaudi 3 AI accelerator will be officially put into production in the second half of 2024. It is reported that this “strongest AI chip” that has attracted much attention before is benchmarked against Nvidia H100. Nvidia said that Gaudi 3 can improve the performance per dollar by about 2 times compared with H100 in terms of reasoning and training.
The first batch of client processors Panther Lake and server processors Clearwater Forest using Intel 18A process nodes are expected to be launched in the market in 2025.
In the long run, it may be a little late for Intel to shout “AI Yes”.
Wells Fargo statistics show that Nvidia currently has a 98% market share in the data center AI market, AMD (AMD.O) has a market share of only 1.2%, and Intel has less than 1%. Bank of America predicts that Gaudi 3 may only occupy less than 1% of the AI ​​accelerator market share in the future.
In AMD’s second quarter report released a few days ago, its data center revenue was about US$2.8 billion, a year-on-year increase of 115%. AMD’s total revenue in the second quarter was US$5.835 billion, less than half of Intel’s revenue this quarter. But AMD’s current market value is more than twice that of Intel.
Intel CEO admitted that the company “has not fully benefited from powerful trends such as artificial intelligence”, resulting in “too high costs and too low returns”.
It is not easy to succeed in the field of artificial intelligence. Zhang Xiaorong, director of the Deep Technology Research Institute, bluntly stated that Intel’s AI chips have insufficient presence in the field of artificial intelligence, weak product competitiveness, and there is indeed room for improvement in technology research and development and market expansion.
Wafer foundry split, frequent losses
Another business line closely related to AI is wafer foundry.
Intel has high hopes for this. Public information shows that since 2023, Intel has successively announced plans to build semiconductor manufacturing plants in many places such as the United States, Europe and Israel, with a total investment of up to US$100 billion.
The performance of the wafer foundry business does not seem to be so amazing. The second quarter financial report showed revenue of US$4.3 billion, a year-on-year increase of 4% and a month-on-month decrease of 1%.
Intel CFO Dave Zinsner pointed out in his speech that the revenue growth of wafer foundry was mainly due to the increase in wafer production of Intel 7 process technology and Intel 3 and Intel 4 process technology. Among them, the latter uses EUV (extreme ultraviolet lithography) nodes from ASML.
However, if the operating loss is taken into account, Intel’s wafer foundry business performance is not optimistic. In the second quarter of this year, the operating loss of the wafer foundry business further expanded to US$2.8 billion, and the operating profit margin was -65.5%.
Intel also admitted that more than 85% of the wafer output still comes from EUV nodes, and the continued expansion of wafers with Intel 3 and Intel 4 process technology in Irish factories has put pressure on the company’s profitability.
This trend will continue in the short term. “We expect operating losses in the third quarter to continue to grow at roughly the same rate.” Dave Zinsner said.
Guo Junli, research director of IDC Asia Pacific, used “difficult” to evaluate Intel’s foundry business in an interview with a reporter from the Times Weekly. “From the input side, wafer fabs need to make large investments in new equipment, plant areas, personnel, etc. From the revenue side, wafer foundry companies TSMC and Samsung have considerable advantages over Intel. As a new entrant, Intel needs to catch up as soon as possible based on technology and customers.” She believes that it may take until 2027 for this part of the business to break even.
Overall, Intel lowered its third-quarter performance guidance expectations, and revenue is expected to be between US$12.5 billion and US$13.5 billion.
The interweaving of excess profits in the distant future and the current risks of change has made Intel’s semiconductor industry face a rare contradiction of talent shortage and layoffs at the same time.
McKinsey data shows that there may be about 70,000 job vacancies in the US semiconductor industry. The PwC report pointed out that by 2030, the talent gap in the European semiconductor industry will reach 350,000. In my country, according to the China Semiconductor Industry Association, the total demand for Chinese industry talents in 2024 will be about 790,000, the talent gap will reach 230,000, and the talent gap in chip design and manufacturing will be about 100,000.
“Companies in the semiconductor industry often span emerging markets and mature markets, and the different demands of these two types of markets lead to this result.” Zhang Xiaorong believes that with the continuous development of new technologies such as artificial intelligence, the demand for semiconductor chips is also increasing. These emerging markets need new products, which has led to a shortage of talent in the semiconductor industry. In mature markets, due to the uncertainty of the global economic situation and the intensification of market competition, the performance of many semiconductor companies has been affected, which has led some companies to choose to lay off employees to reduce operational pressure.
