In a recent statement, Intel's CEO Pat Gelsinger announced that Intel will completely spin off its chip foundry business to become an independent subsidiary and will introduce external financing.
This decision is seen as a significant departure from Intel's IDM (Integrated Design and Manufacturing) model, which it has adhered to for over fifty years.
The IDM model means that companies maintain independent control over design, production, packaging, and other stages, achieving complete integration of the supply chain.
In stark contrast, companies like Nvidia, AMD, and Qualcomm adopt a Fabless model, only handling design, while outsourcing other manufacturing and packaging stages to specialized chip producers like TSMC and Samsung.
Intel's abandonment of the IDM model it has followed for fifty years raises the question: Is this a path to self-redemption, or is it self-sabotage?
How will this move shake the global chip market? What does it mean for us?
In fact, Intel's spin-off of its chip manufacturing business had long been anticipated within the industry.
In the first two quarters of this year, Intel faced enormous losses, with the second quarter alone seeing a staggering loss of $1.6 billion.
During that quarter, the chip manufacturing division alone incurred losses of $2.8 billion, nearly overshadowing Intel's growth in data center and artificial intelligence sectors.
Following the earnings report, Intel's stock price plummeted over 30%, with a cumulative decline of 56% this year.
Faced with such a dire situation, Intel announced a massive layoff in early August, affecting over 15,000 employees.
Speculation surrounding its spin-off of the chip foundry business became rampant, and now it has finally come to fruition.
It's ironic that what was once a source of pride for Intel, its chip manufacturing plant, has now become its biggest burden.
At its peak, Intel commanded over 85% of the CPU market share with its advanced process technology, almost achieving a monopoly.
In 2014, Intel's 14nm process was hailed as the world's leading technology, even comparable to TSMC's 16nm and 20nm processes, asserting that it was far ahead in chip manufacturing.
Back then, Intel completely rejected external chip foundry orders, even high-priced offers.
Today, however, Intel sees its chip manufacturing division as its biggest liability.
What caused such a drastic shift in Intel's attitude towards its chip manufacturing division in just a decade?
The fundamental reason lies in the gradual decline of the American chip manufacturing industry.
Since Intel touted its 14nm process as globally leading in 2014, its chip manufacturing capabilities have begun to decline.
During the 10nm process stage, Intel failed to achieve a technological breakthrough, instead relying on its market monopoly to continue slowly progressing, leading outsiders to mockingly call it the "toothpaste factory."
At the same time, TSMC made continuous advancements in 7nm, 5nm, and even 3nm process technologies, gradually overtaking Intel, which felt the pressure.
In 2021, Intel appointed a new CEO and began aggressively pushing process development, planning to leap five process nodes within four years to catch up with TSMC's nearly decade-long progress.
However, by then, the American chip manufacturing industry had long lost its former glory.
Despite Intel investing over $25 billion annually in manufacturing projects, its aggressive policies led to severe quality issues in high-end CPUs.
In May this year, a reviewer pointed out that Intel's latest two generations of high-end CPUs had defects, with the oxidation-resistant coating inadequately applied, resulting in severe issues like crashes and blue screens under high-pressure conditions.
This news not only caused another drop in stock price but also eroded global user confidence in Intel's products, leading to a continuous decline in market share.
Interestingly, TSMC, Intel's biggest competitor in the chip foundry sector today, was somewhat nurtured by Intel itself.
Back when Morris Chang envisioned the chip foundry business, he invited Intel to collaborate, but Intel arrogantly believed that only the U.S. could produce the most advanced chips.
Now, even Intel can hardly bear the enormous losses in American chip manufacturing, choosing to spin off its own chip foundry division.
This is undoubtedly bad news for the U.S. government. Intel, as a representative of American chip manufacturing, was once called the "American champion semiconductor company" by the Commerce Secretary, and its spin-off of the foundry business signifies a further weakening of its competitiveness.
Reflecting on AMD's experience, in 2008, due to consecutive losses, AMD split its chip foundry into an independent subsidiary, GlobalFoundries.
Although AMD's decision was once viewed as a departure, the later GlobalFoundries ultimately abandoned processes below 7nm due to an inability to participate in costly advanced process R&D, leading to a shrinking market share.
According to data from the second quarter of this year, GlobalFoundries' market share has dwindled to just 5%, even less than SMIC, which faces numerous restrictions.
This phenomenon precisely reflects the changes in American manufacturing, with rising raw material and labor costs, coupled with executive compensation highly linked to stock prices, leading to relatively reduced funding for R&D, ultimately resulting in declining product quality and frequent issues.
In such circumstances, executives often choose to divest loss-making divisions to achieve a short-term rebound in stock prices.
The spun-off factories may struggle to survive like GlobalFoundries or face closure.
Intel's recent decisions bear a striking resemblance to the choices made by past companies like AMD and Boeing.
After experiencing losses and stock price pressure, Intel similarly opted to divest its manufacturing division.
Even so, whether the newly formed subsidiary can achieve greater development remains uncertain.
The investment cost required for each advancement in chip processes has significantly increased, and whether external funding can be effectively injected is still a question.
At the same time, Intel recently announced that it would suspend plans for new factories in Germany and Poland and consider retracting its existing packaging and testing operations from Malaysia.
This will significantly reduce support for the chip manufacturing division while also indicating further cost increases.
Additionally, although TSMC and Samsung's new factories in the U.S. are expected to start production before the end of 2025, Intel will no longer be the sole advanced process chip producer in the U.S., further diminishing its bargaining power.
In the context of the U.S.-China tech war, America's blockade against China's chip technology increasingly appears ineffective.
As China catches up in advanced manufacturing capabilities, the competition Intel faces will come not only from TSMC.
Finally, Israel's recent terrorist activities in Lebanon are bound to have a negative impact on the Western electronics industry.
Given Intel's currently poor product reputation, whether its chip foundry business can secure orders in the international market remains an unknown.
Comments