There is no simple answer, but we think there are a few factors that really stand out...:
By now, we are all familiar with the fact that TSMC is, by far, the most capable semiconductor manufacturer in the world, with all the entails for the industry and geopolitics. And as this reality sets in, many people have been asking us how did they get so good?
There are many good histories of this process, the most recent of which is Chris Miller's Chip Wars (which is very good), that cover the narrative of TSMC's rise, but we think the question really goes to deeper fundamentals. What capabilities does TSMC have that others lack? And can anyone else build those? Not surprisingly, there is no simple answer to this question, but we think there are a few factors that really stand out.
[...] It is important to understand that much of TSMC's current advantage rests in intangibles. Everyone else can buy the same equipment. And their success is not the result of one genius. China's SMIC has poached hundreds of TSMC staff over the years, to no avail. Nor is the answer capital -- building an advanced fab is incredibly expensive, but many countries have sunk vast fortunes into futile attempts.
TSMC's advantage rests in years of learning and process development. They know how to prepare chip designs for production, build them, and most importantly fix the inevitable bugs that appear. Most importantly, they are able to capture and retain all this learning, so they can do all of this at immense scale, repeatedly.
[...] Finally, and perhaps most importantly, TSMC produces more silicon wafers than anyone else, by a considerable margin. Practice leads to improvement and TSMC gets more practice than anyone else. It may seem counter-intuitive, but semis manufacturing is as much art as science.
[...] All of this provides TSMC not only with important advantages in capabilities, but also in terms of economics. There are plenty of places in the world with the ability to produce chips on leading edge processes, and give those academics enough time they can probably yield a perfect wafer, but TSMC can get there faster and in massive volume.
All of this is to say, that it is unlikely that anyone can truly catch up with TSMC anytime soon. It took them decades to build up these processes. And this helps partially explain why TSMC is now willing to build advanced fabs outside Taiwan. They know that their fabs in Arizona and Japan and Germany and wherever are going to remain dependent on the company's deep expertise back in Taiwan. The US may get a leading edge fab, but that does not mean the US can really do what TSMC does.
(Score: 1, Insightful) by Anonymous Coward on Friday December 23, @08:46AM
Even if the first claim is true, the latter statement doesn't actually prove the first claim.
For example if you poach many waiters, dishwashers, assistant chefs etc from a restaurant but you don't poach the head-chef and his secret recipes; it doesn't prove that the restaurant's greatness is not due to a single genius.
(Score: 2, Funny) by dwilson98052 on Friday December 23, @09:25AM
...but it was aliens.
(Score: 1, Interesting) by Anonymous Coward on Friday December 23, @01:10PM (1 child)
The following is based on reading over the years, and also conversations with a friend who worked on process control at an (un-named) foundry.
Intel historically used a different process than everyone else to open a new chip plant(s). First they would get the yield up for a new technology (smaller size features) at one plant--heavy with scientists, engineers and techs. Then they would move heaven and earth to copy _everything_ about that plant to the new plant(s)--not just choice of equipment, but the supply chain, floor plan, clean room design, everything. If everything cloned near-perfectly, the new plant could operate without all the "brains" at the original fab. When it works, this might be the cheapest way to ramp up volume? But iirc, it didn't always work, as feature size shrunk, it got harder and harder to duplicate conditions from one site to another.
Just about every other foundry had smart/clever people locally. Cost more for the staff, but when there were problems they could be solved locally, much quicker than flying in the experts from the corporate brain trust. This also means that innovation could happen in more than one team, potentially giving an edge in solving problems across the whole company.
How does TSMC approach this problem?
Can remote access and web-conferencing replace having the best corporate problem solvers on-site--or do you really need to be there in person?
(Score: 0) by Anonymous Coward on Saturday December 24, @10:55AM
More managers, longer meetings, progress reports, short-term contracts, job insecurity. Get people motivated by existential fear.
That's got to be it, ain't it?
(Score: 4, Interesting) by RamiK on Friday December 23, @01:14PM
The Taiwanese stock exchange opened to foreign investors in the early-mid 80s.
The US-Japanese semiconductors accords were signed around '86 to stop Japan's semiconductor industry from taking over.
So, when Morris got back home from TI in '87 just as Taiwan became a democracy and started up the first fabless business model in the world, the money was flowing to the degree teens were speculating in stocks after school as a pastime ( https://www.asiancenturystocks.com/p/boombusttaiwan [asiancenturystocks.com] ) and there was this nice big vacuum the Japanese left over just waiting for TSMC's taking...
Anyhow, Taiwan was also diversify out of its drying mining industry into biotech but the returns there just couldn't match anything like what the Japanese withdraw from the markets gave.
Anyhow, Israel and mainland China also emerged around those years and more than a few European companies got their breaks too but the money just wasn't flowing to tech as fast as it was in Taiwan.
compiling...
(Score: 4, Interesting) by Rich on Friday December 23, @02:24PM (1 child)
There's a channel on YT, "Asianometry", which reports about the semiconductor industry and its history, with a special focus on Taiwan. What I took away from there is that Intel stumbled trying to introduce a 153nm wavelength laser process, while TSMC stayed at 193nm and introduced immersion lithography. And there was another large Taiwanese shop, UMC, which lost its contact to the leading edge over a process sequence detail ("gate first" vs. "gate last") and business decisions where TSMC was at an advantage because of its pure "fab-only" model.
The channel also tells a story how Samsung got to the leading edge with the help of a disgruntled TSMC guy (Liang Mong Song?!). All quite interesting, although I'm not sure I could fully follow on non-semiconductor topics like the assessment of AGR power plant performance in the UK.
(Score: 0) by Anonymous Coward on Friday December 23, @10:19PM
I watch the Asianometry channel, he does a good job covering semiconductors (and many other cool topics). Highly recommended.
On-topic of TSMC, I've heard one of the many factors that isn't commonly cited is increased inspection. They have more equipment checking for defects than their peers, which implies less defects and tighter control. This also matches their customer base: if you are selling specific purpose chips they need to work to spec. This is in contrast to AMD/Intel, where they can tolerate a higher percentage of sub-optimal chips through chip binning. At first, being able to recover a potential i9 chip as an i5 is a boost to profitability since you aren't scrapping that chip. But it can lull the company into poorer process control overall.
(Score: 3, Funny) by Mojibake Tengu on Friday December 23, @06:28PM
Of all over the world semiconductor manufacturers, TSMC offices and facilities have best horoscopes and Feng Shui setup.
It's simple as that.
Also, never underestimate the horoscopes of your personnel or business partners.
The edge of 太玄 cannot be defined, for it is beyond every aspect of design