Can Samsung surpass TSMC within 10 years? Expert reveals the gap with one phrase: 'Don't eat from your bowl while eyeing others'

South Korea's Samsung achieved historic performance in the first half, benefiting from the artificial intelligence (AI) demand surge. However, TSMC Chairman Mark Liu stated at the early June shareholder meeting: 'Twenty years ago, they said they'd catch up to TSMC in 10 years. Ten years ago, they said the same. Recently, they again claim they'll catch up in 10 years. My comment? They're dreaming.'

In response, veteran tech journalist Lin Hongwen analyzed on the program 'How Bojie Sees It,' stating that many people think about 'eating from their bowl while eyeing others'—but this isn't necessarily right.

Lin explained that Samsung's market is memory, so TSMC has never claimed to surpass Samsung in that domain. Former TSMC Chairman Morris Chang foresaw this trend and in 2000 announced that its subsidiary 'Vanguard International Semiconductor' would exit dynamic random-access memory (DRAM) manufacturing. As Lin noted, 'Catching up to them (Samsung) is too difficult.'

Lin further pointed out that South Korea's government announced in late June a major investment to build semiconductor production bases and AI data centers. However, the investment plan 'excludes foundry services and integrated circuit (IC) design.' Lin believes this shows South Korea aims to solidify its existing market rather than 'take others' food.'

Lin emphasized that as long as South Korea strengthens its own market, it holds infinite opportunities. 'Many people think about “eating from their bowl while eyeing others,” but that's not necessarily correct,' he said.

Lin noted that when TSMC was founded in 1987, 'who could have known foundry services would become such a huge market?' Yet TSMC has now been in this business for nearly 40 years. 'Can you imagine the immense value it has created?' Lin stressed that companies should focus their attention on what they do best within their capabilities.

Lin also urged the United States to concentrate on application fields and standard-setting, and to stop manufacturing, as Asia excels in production. He predicted that the U.S. 'Tera Fab' project—a super semiconductor manufacturing plant jointly pushed by Tesla, SpaceX, and xAI—will not succeed.