Intel’s Lost Decade: How TSMC Took the Chip Lead
Intel’s Lost Decade: How TSMC Took the Chip Lead
In the Wall Street Journal on Saturday, December 7th, the lead story on Intel was trying to place the blame on outgoing CEO Pat Gelsinger. The demise of Intel started years before Gelsinger returned in 2021.
This blog looks back at the last ten years and looks at how Intel fell behind TSMC in manufacturing advanced chips. We’ll explore the reasons for this shift and the impact it has had on the market.
Why Did Intel Fall Behind TSMC?
For decades, Intel was the undisputed leader in chip manufacturing. However, over the last ten years, Taiwan Semiconductor Manufacturing Company (TSMC) has emerged as the new leader, particularly in thinner and thinner processors. This shift can be attributed to several factors but most of them were issues that Intel caused itself:
- Delay after Delay The delays in launching new and thinner chip architectures started as far back as 2014. In those days Intel took risks but its own launch delays was what contributed to the slow rise of TSMC.
- In 2015, When Brian Krzanich was Intel CEO – Intel’s 10 nm process was delayed, meaning that its 14 nm processors had to live on for longer the average 2 years. The issue that Intel admitted to was under investment in EUV lithography machines that were made by ASML (Ironically under Gelsinger Intel recently invested in huge purchases of ASML machines).
- In 2020, Intel missed its deadline for the launch of its 7 nm processor, which cleared the way for Gelsinger’s return.
- Missed Opportunities: Intel missed key opportunities in the mobile market, focusing on high-margin PC and server chips. This allowed TSMC to gain significant volume and experience through partnerships with mobile chip designers like Apple. In fact, Apple signed a deal with TSMC in 2013 for it to start producing chips for the iPhone A8 processor.
- Technological Choices: Intel’s reliance on older manufacturing techniques, like deep ultraviolet (DUV) lithography, delayed its development of advanced nodes. Meanwhile, TSMC’s early adoption of extreme ultraviolet (EUV) lithography gave it a crucial advantage.
- The Rise of Fabless Companies: TSMC’s pure-play foundry model enabled the growth of fabless chip companies like NVIDIA and AMD. These companies, unburdened by the cost of manufacturing, could focus on design and innovation, further driving TSMC’s growth.
Analysis
Aragon Research’s take is that you can’t place the blame on Pat Gelsinger. He only came back to Intel in 2021. The seeds of Intel’s demise started long ago. Intel’s loss of leadership has had significant consequences. AMD, once a struggling competitor, has gained market share with its TSMC-manufactured chips that outperform Intel’s offerings. Apple’s adoption of its own TSMC-manufactured M1 chips has further highlighted Intel’s struggles.
Intel is Not Changing its Strategy that Gelsinger Developed
In all of the articles published in the last few days, Intel executives admitted that the Intel strategy will most likely not change. However, we would suggest that fundamental issues remain with Intel – and that until the right team is put in place – delays due to multiple issues – will not go away.
TSMC’s dominance has empowered fabless chip companies, leading to increased competition and innovation. This trend is likely to continue, with TSMC’s advanced nodes driving the development of new and more powerful technologies. We would note that Intel is pushing for a 1 nm process – to try to regain technological leadership.
Bottom Line
Intel’s failure to adapt to the changing semiconductor landscape has allowed TSMC to take the lead in chip manufacturing, particularly for some of their former customers, including Apple.. This shift has significant implications for the entire technology industry. Enterprises should stay informed about these developments and when investing in data centers, consider the processor architectures being used, the refresh rates of those processors and who actually manufactures them..
Have a Comment on this?