That’s a interesting point and it does make some sense. However, I can see how others—particularly many in China—might argue that SMIC was already on a successful trajectory, and that the only major obstacle to its continued rise has been the EUV embargo.
Ex-Intel CEO backs Trump's sovereign wealth fund plan — and says it can help the US keep up with China in the tech race
- Thread starter Daniel Nenni
- Start date
That’s a interesting point and it does make some sense. However, I can see how others—particularly many in China—might argue that SMIC was already on a successful trajectory, and that the only major obstacle to its continued rise has been the EUV embargo.
FWIW.. According to wikipedia: Although the government of Taiwan is the largest individual shareholder,[11] the majority of TSMC is owned by foreign investors
i.e. Government involvement is substantial.
Also:
When Chang founded TSMC in 1987, he did so with $100 million in seed money from the Taiwanese government.
TSMC and Taiwan’s Government: Two Boats on the Same Tide - Domino Theory
The simple answer is an alignment between the interests and constraints of TSMC and the Taiwanese government. The government was instrumental in TSMC’s founding. Since then, a mutual dependency has developed. While the Taiwanese government may feel some apprehension about TSMC’s current outbound...
dominotheory.com
TSMC certainly succeeded because it was private and well run; but it wouldn't have existed without government.
A sovereign wealth "bad bank" that can buy bankrupt chip companies, restructure them, (looking at you Wolfspeed) and have the patience, like with Fanny Mae and Freddy Mac, to just wait 15 years for profitability to return, that is something the US knows how to do and succeed at.
hist78
Well-known member
Sounds like subsidized wafer yield lossBiggest incentive to not improve.
I remember this comic:
Are you kidding me ？
If this can turn a profit, capital will move in.
The biggest problem with government intervention isn’t that it’s guaranteed to lose money (I’d bet it will), but that no private investor will ever want to put money here again.
I know it so well.
I get your point that the CCP capital brings CCP objectives to achieve certain goals other than profitability drives out private capital from the Chinese chip market. That's fair.
How you avoid it is, when a firm goes bankrupt, private capital is wiped out. Restructured to zero. Shares have no value.
Then the government steps in, in some manner. The FDIC resolution of Silicon Valley Bank. The Resolution Trust Company during the 1980s Savings and Loan crisis. Or in 2009, the conservatorship of Fanny Mae. There are different names but the same basic process. No bail out. The government has skin in the game but generally behaves passively. And no private equity draining the firm like a vampire.
haha, funny story。
SMIC was stuck at 28nm for many years.
It was only after Liang Mong-Song’s arrival that SMIC first cracked 14 nm, then 7 nm, and finally went back to re-tune 28 nm—at which point the 28 nm node could be called a volume-production success.
SMIC’s success was more to luck than skill.
FWIW.. According to wikipedia: Although the government of Taiwan is the largest individual shareholder,[11] the majority of TSMC is owned by foreign investors
i.e. Government involvement is substantial.
Also:
When Chang founded TSMC in 1987, he did so with $100 million in seed money from the Taiwanese government.
TSMC certainly succeeded because it was private and well run; but it wouldn't have existed without government.TSMC and Taiwan’s Government: Two Boats on the Same Tide - Domino Theory
The simple answer is an alignment between the interests and constraints of TSMC and the Taiwanese government. The government was instrumental in TSMC’s founding. Since then, a mutual dependency has developed. While the Taiwanese government may feel some apprehension about TSMC’s current outbound...
I would argue that TSMC would not have succeeded without Phillips. Phillips made a big investment (27% stake) and made the technology available for TSMC to get started. Phillips also gave TSMC a global reach and the confidence fabless customers needed to take the initial risk with a new foundry based in Taiwan. Morris Change managed the relationship with Phillips and the Taiwanese Government so he deserves the credit for making it all happen.
Your article on this topic (I just read) is quite good - https://semiwiki.com/semiconductor-manufacturers/333584-how-philips-saved-tsmc/ . I had not realized how critical they were.
Still, it sounds like the Taiwanese government put up even more money (48%) so it probably required both to succeed? Though I know the overall global capitalization situation was greatly improved in the mid-80s vs. the prior decade, maybe they could have found other sources.
without Philips and Morris, TSMC will certainly not start the way they did. But given the government focus and conviction, it is probably a matter of time.
hist78
Well-known member
Philips' $40 million initial investment in TSMC, about 27.6% of TSMC’s startup capital, was critical. But there are nuances to understand.
- 1. During the founding negotiation, Philips demanded TSMC fully follow and use Philips' semiconductor manufacturing technology. This was rejected and not included in the final agreement. Morris Chang explained that he felt a successful foundry must have its own technological capabilities and R&D independence. TSMC ended up picking and choosing what it needed to improve its own manufacturing process and agreed to pay Philips a 3% royalty (based on revenue) for products under 1 micron. TSMC also rejected Philips’ proposal to exchange technology licensing for TSMC shares.
- 3. One obvious reason why TSMC did not want to fully copy Philips' technology is that Taiwan had been working on semiconductor manufacturing since the early 1970s. Although still small, Taiwan already had a solid base to move forward by the time TSMC started business in 1987. Some milestones include:
- 1973:
The Industrial Technology Research Institute (ITRI) was founded as a technology R&D institution in Hsinchu City, Taiwan.
- 1976:
ITRI signed a technology transfer agreement with RCA, specifically for CMOS IC technology.
- 1976–1977:
ITRI sent engineers to RCA for training and built a demonstration factory in Hsinchu City. The fab achieved a 70% yield rate, surpassing RCA’s original 50% yield rate. (Today, in 2025, we all know how important the yield rate is in the semiconductor industry.)
- 1980:
United Microelectronics Corporation (UMC), Taiwan's first IC IDM company, was spun off from ITRI, taking over the 4-inch wafer technology.
- 1984:
ITRI launched the VLSI project to develop 6-inch wafer technology and boost semiconductor manufacturing.
- 1987:
ITRI spun off a chip manufacturing facility, transferring fabs, equipment, technologies, and 98 professionals to TSMC.
By the time TSMC started in 1987, Taiwan was probably only 2 to 3 generations behind the world’s most advanced process technologies. If we compare TSMC's start with India’s new semiconductor startups or Japan’s Rapidus venture, the difference is dramatic. India is too far behind in semiconductor manufacturing, and Rapidus started from scratch and adopted IBM's 2nm technology,
- 4. Philips’ $40 million (27.6%) investment in TSMC was most impactful in lending its credibility to support the Taiwan (ROC) government’s 48.3% stake and in persuading Taiwanese private investors to fill the remaining 24.1% of capital. Taiwan government did not want to increase their share at TSMC to more than 50% because it would make TSMC as a government owned business.
5. We also have to recognize that Intel outsourcing some product manufacturing to TSMC in 1988 was a very important event. In Morris Chang’s own words, Philips was in the front row of second tier semiconductor manufacturing, while Intel was the undisputed number one in the industry back then. Intel’s orders gave TSMC a huge boost in credibility.
Source: ITRI, Morris Chang Autobiography Part II, Google.
Last edited: