The Complete History & Strategy of TSMC
Podcasts have increasingly become the primary source of learning for many, partly because of the convenience of listening to it while on a run or in the gym or on commute as opposed to reading. This podcast by Ben and David in particular is very extensive in terms of background research on the topic of the podcast. In this episode, they discuss the history and strategy of TSMC or Taiwan Semiconductor Manufacturing Company – the world’s largest semiconductor manufacturing company and the ninth most valuable company in the world. Given the ubiquity of semiconductors in our lives, understanding the business from the perspective of the largest company to make them is not just interesting but also from the perspective of understanding utterly dominant companies (central to Marcellus’ investment philosophy) – 90% market share in the leading edge 5 nanometre chips. “They make literally every chip in every iPhone sold today and soon to be in every Mac sold. If you’re excited at all about Nvidia, AMD, Qualcomm, or even any of the chips that Amazon, Microsoft, Facebook, and Apple are making—all of those chips or nearly all of them are actually made by TSMC. Along with all the chips in your cars, in your smart home devices, fighter jets, and everything.”
Ben and David discuss the founder Morris Chang’s own background to begin with and how he actually founded TSMC by order of the Taiwanese govt when he was in his late 50s after a successful career in the US semiconductor industry, was in a semi-retired job leading a govt run technology research outfit in Taiwan. The back story has some interesting tidbits about how Morris started his stint at Texas Instruments (TI) the then largest semiconductor company in 1958, the year in which the Integrated Circuit (IC) was invented, which eventually disrupted the field of electronics and how TI lost out to Intel in the race to make the chip for the IBM PC.
They attribute TSMC’s success to its decision to focus on one and only one thing – foundry and rather partner with everyone else in the value chain than try and vertically integrate themselves.
But the decision was also driven of the situation in Taiwan back in the 80s.
“We had no strength in research and development, orvery little anyway. We had no strength in circuit design, IC product design. We had little strength in sales and marketing,” and this is of course, referring to Taiwan as a nation, “and we had almost no strength in intellectual property. The only possible strength in Taiwan that we had, and even that was just a potential one, not an obvious one, was semiconductor manufacturing, wafer manufacturing. And so what kind of company would you create to fit that strength and avoid all the other weaknesses? The answer was a pure-play foundry.”
….The Taiwanese government wanted to come up in the world. They’re like, this is where we are now, this is not where we want to be. They knew that technology was the way. They had decided back in the ‘70s that they would establish an initiative called the Industrial Technology Research Institute or ITRI, and the goal was for it to become like the Bell Labs of Taiwan to do some tech transfers from the US and elsewhere, home grow some real technology businesses in Taiwan so that maybe they can lift businesses out of poverty there at least.”
TSMC was the first pure-play foundry when all semiconductor companies insisted on making their own chips i.e, there was no market for a pure-play foundry. As it turns out, eventually almost all semiconductor companies became fabless creating a huge market for TSMC.
“When I was at TI and General Instruments I saw a lot of integrated circuit designers wanting to leave and set up their own business, but the one thing or the biggest thing that stopped them from leaving those companies was they couldn’t raise enough money to form their own company. Because at the time,” as we were just saying—real men, “it was thought that every company needed manufacturing, needed their wafer manufacturing, and the most capital intensive part of a semiconductor company, of an IC company (does the manufacturing). I saw those people wanting to leave, but being stopped by the lack of ability to raise a lot of money and build a wafer fab.”
Also fascinating to know that Morris Chang received no equity at the get-go (TSMC was 50% owned by the Taiwanese govt, 28% by Philips and the rest by other Taiwanese companies which were strong armed by the premier to put up capital – TSMC shares have now compounded at 20% for 27 years since listing). Morris had to build his stake (worth $3bn now) by buying shares from his own savings.
Ben and David then discuss the whole semiconductor value chain – a tutorial of sorts, very important to understand how TSMC’s dominance is secured. An important reason being another utterly dominant company in the value chain – ASML, that produces critical high end lithography machines for the semiconductor industry, the only one in the world to do so. ASML is a Dutch company and the Netherlands haven’t renewed their trade agreement with China – so that blocks out any competition from China to TSMC. The criticality and complexity of what ASML does in the semi business:
“they cost $200 million for a machine that makes the chips…. It takes four 747s to ship one of these machines. You’re TSMC, you buy one, and it arrives. Of course, the 747s, then there’s a crew of ASML employees on-site, not only to assemble it, but then to help you run it. Like you mentioned, these companies are deeply integrated with each other to be able to pull this off.
It becomes exponentially harder to manufacture chips the more dense they are. So David, you mentioned 150 nanometers or so from several years back. We know now that the M1s are made on this 5-nanometerprocess. Well, the wavelength of white light (of regular light) is 193 nanometers.
…You have to create a plasma. What they do, and this is called extreme ultraviolet light or EUV—this is a process that is just wild. On one side of the machine, you drop molten tin. On the other side of the machine, you then hit it with a highly specialized laser. You perfectly pulse them. It explodes into a plasma, which creates extreme ultraviolet light. Of course, this is hard enough to do as you can imagine how that might work, but you actually have to do that 50,000 times per second. What I read is that the accuracy with which that laser needs to hit the drop of the molten tin is more precise than the calculations to send the Apollo missions to the moon. You got to do that 50,000 times a second.”
There are only so many machines that can be made in a year. And TSMC has ordered everything in ASML’s orderbook.
There is also a fascinating section on how Morris came back from retirement at the age of 78 to strike the Apple deal – arguably the most valuable partnership in business history.