"You Don't Need It, Until You Need It," Says Analyst on Whether Intel Could Replace TSMC
Intel’s position in the semiconductor industry has shifted significantly over the past few years, especially as the United States pushes to strengthen domestic chip production. Once focused primarily on its own processors, Intel is now evolving into a potential safety net for the U.S. chip ecosystem. The question, however, is whether Intel can realistically serve as a replacement for TSMC, the world’s leading chip manufacturer.
According to analyst Ben Bajarin, who discussed this topic in an interview with Stratechery (via Ray Wang on X), Intel’s existence is crucial for America’s fabless semiconductor companies such as Apple, AMD, NVIDIA, and Qualcomm.
“At the end of the day, Intel existing is the single most important piece of insurance for an Apple, for an AMD, for an Nvidia, for a Qualcomm, for basically every single U.S. fabless chip company. The reason why insurance is the right analogy is because in any normal state of affairs, they don’t need Intel.”
Bajarin explained that while TSMC currently provides the best manufacturing capabilities in the world, Intel represents an insurance policy in case geopolitical tensions disrupt supply chains. As he put it, “You don’t need it until you need it.”
Currently, Taiwan produces over 90% of the world’s advanced semiconductors, according to U.S. Commerce Secretary Howard Lutnick, with TSMC contributing the majority of that output. This concentration creates significant global risk. If Taiwan’s chip production were ever interrupted, the entire technology industry would face severe shortages of advanced processors.
To reduce this dependency, TSMC has accelerated its U.S. expansion plans, with over $300 billion in investments aimed at establishing leading-edge manufacturing facilities in America. This includes bringing its 2nm N2 process online earlier than expected. However, shifting production from Asia to the U.S. is a long-term project that will take years to complete and will require extensive political and financial backing.
Meanwhile, Intel Foundry Services (IFS) is positioning itself as the most viable U.S.-based alternative. To reach TSMC’s level, Intel must deliver competitive production yields, capacity, and pricing with its upcoming Intel 18A and Intel 14A nodes. If these nodes succeed, Intel could attract more fabless clients and reestablish process leadership in the U.S. semiconductor market.
Industry veterans such as Jim Keller agree that Intel’s intention to rebuild its foundry business is clear, but the company still faces major challenges. Competing with TSMC requires not only advanced technology but also global trust and a consistent track record of execution.
Ultimately, Intel’s potential as a TSMC alternative is about more than market competition. It represents industrial resilience and national security for the U.S. As Bajarin stated, Intel may not be essential today, but in times of crisis, its existence could safeguard the entire American tech industry.
What do you think? Can Intel truly step up as a reliable alternative to TSMC, or will TSMC’s dominance continue unchallenged?