Recently, at the Playground Global conference in Tokyo, Japan, former Intel CEO Pat Gelsinger publicly discussed for the first time the reasons for his departure in December 2024. He stated that the decision to resign was made by a “third party” rather than his personal will, and that he was unable to complete the IDM 2.0 strategy he had been pushing at Intel. Gelsinger led Intel for over four years, with his core goal being to transform Intel from a traditional chip design and manufacturing company into a leading global integrated foundry service provider, in response to challenges from competitors like TSMC and Samsung. However, the continuous losses in the foundry business, insufficient competitiveness in the end-product market, and internal strategic disagreements led to his early departure. After the new CEO, Kevin Chen, took office, Intel is adjusting its strategic direction, focusing on 18A process mass production, expanding external cooperation, and optimizing its organization, attempting to reshape its market position.
Gelsinger returned to Intel as CEO in 2021, proposing the IDM 2.0 strategy, hoping to revitalize Intel’s competitiveness in the semiconductor industry by integrating three pillars: internal manufacturing, external foundry cooperation, and its own foundry services. The core of this strategy was the “5N4Y” plan, which aimed to launch five advanced process nodes within four years, including 10nm, 7nm, 4nm, 3nm, and the more advanced 18A (1.8nm) and 14A (1.4nm). Intel hoped to catch up with TSMC’s 2nm process and become the world’s second-largest foundry by 2030. During the implementation of IDM 2.0, Intel’s foundry business faced multiple challenges. In 2023, Intel Foundry Services (IFS) reported a staggering loss of $7 billion, far exceeding expectations, with limited customer orders and a lack of sufficient market appeal for the 18A process. In contrast, TSMC’s 2023 revenue was approximately $69.3 billion, accounting for about 60% of the global foundry market, with Samsung close behind, while Intel’s market share was less than 10%.
During Gelsinger’s tenure, Intel made some progress in process technology. For example, Intel 4 (4nm) has been used in Meteor Lake processors, Intel 3 is being prepared for data center chips, and the 18A process is expected to enter mass production in 2025, planned for Panther Lake client processors and Clearwater Forest server chips. Intel also reached an agreement with Microsoft to customize 18A chips for them, showing some market potential. However, several process nodes were limited to internal use due to performance or cost issues and failed to attract external customers. In addition, although Intel innovated in advanced packaging technologies (such as Foveros and EMIB), it lagged behind TSMC in EUV lithography machine application and yield optimization, leading to high production costs. These technical bottlenecks and financial pressures weakened the effectiveness of IDM 2.0 implementation.
The disagreement between Intel’s board of directors and Gelsinger likely stemmed from the long-term losses of the foundry business and the speed of strategic execution. According to industry analysis, the board hoped to quickly improve financial conditions and boost stock prices through more pragmatic strategies. In 2024, Intel’s stock price fell by about 40%, and its market capitalization shrank to approximately $100 billion, far below TSMC’s nearly $900 billion and AMD’s approximately $250 billion. The large-scale investments pushed by Gelsinger, including the construction or expansion of wafer fabs in the United States, Germany, Israel, and other places, costing tens of billions of dollars, further exacerbated the financial burden. Although these investments received approximately $8.5 billion in subsidies from the US CHIPS and Science Act, it will be difficult to reverse the losses in the short term.
The new CEO, Kevin Chen, took office in December 2024. His background includes leadership experience at Cadence Design Systems and a deep insight into the semiconductor industry. Chen quickly took action after assuming office, announcing layoffs of approximately 15,000 people, or 15% of the total workforce, to cut operating costs. In addition, Intel plans to divest non-core businesses, such as the programmable chip division Altera, and suspend some new factory construction projects, for example, the groundbreaking of the Magdeburg factory in Germany has been postponed to 2026. Chen also explicitly stated that he would expand cooperation with external foundries such as TSMC and Samsung to produce some Arrow Lake and Lunar Lake processors, in order to reduce costs and enhance market flexibility.
At the technical level, Intel is accelerating the preparation for mass production of the 18A process. 18A adopts GAA (Gate-All-Around) architecture and PowerVia backside power delivery technology, which theoretically can improve performance-per-watt by 10-15%, approaching the level of TSMC’s 2nm process. Intel expects to achieve breakeven for its foundry business in 2027 and plans to launch the 14A process in 2026, further narrowing the technological gap with competitors. At the same time, Intel is increasing its investment in AI chips; the Gaudi 3 accelerator was released in 2024, claiming superior performance in some scenarios compared to Nvidia’s H100, attempting to gain a share in the AI computing market.
The external environment facing Intel is equally complex. The global semiconductor industry is experiencing demand fluctuations; the PC market is expected to grow by only 3-5% in 2025, and server chip demand has rebounded due to the AI boom, but competition is intensifying. TSMC and Samsung, with their mature 3nm and 2nm processes, have attracted major customers such as Apple and Nvidia, while Intel needs to prove the reliability and cost advantage of its foundry services in a short period. In addition, the US government’s requirements for localization of the semiconductor supply chain bring policy dividends to Intel, but also come with high compliance costs.
Gelsinger’s departure may mark a turning point in Intel’s transformation journey. His IDM 2.0 strategy laid the technical foundation for Intel, but failed to quickly translate into market competitiveness. The new leadership’s adjusted strategy shows that Intel is shifting from aggressive expansion to pragmatic optimization, in an attempt to gain a foothold in the global foundry market. In the coming years, the mass production performance of the 18A process, the expansion of external customers, and breakthroughs in the AI chip market will be key to Intel’s resurgence.