You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!
"CNBC first visited the fab in 2021, not long after TSMC broke ground. TSMC initially announced the plant would cost $12 billion and pump out 5-nanometer chips by the end of 2024. Three years later, that price tag has soared to $20 billion and full production is delayed until 2025.
Instead, the fab is in pilot production, making sample wafers and sending them to customers for verification. TSMC has committed to building two more fabs on the site by the end of the decade, for a total investment of $65 billion.
The project is "dang near back on the original schedule," TSMC Chariman Rick Cassidy told CNBC during an exclusive first look at the completed fab in November."
I wish they supplied more details why the $12B budget ended up $20B.
And why they called out MEP (Mechanical Electrical and Plumbing) as the "biggest challenge that we had"
This sounds standard for large mega projects of which I have worked on more than a few. Tackling problems on very large multibillion dollar projects was part of my job. It's surprising when a mega project doesn't run into a few problems along the way. A billion here a billion there is how these mega projects go. Major trouble shooting was part of my job. I hold three contractor's licenses in California.
One thing lost in the shuffle with the TSMC Arizona fab news is that Taiwan passed a law recently requiring all leading edge semiconductor fabricaton performed by any Taiwan company (basically just TSMC) be done only in Taiwan. So, despite what TSMC told the CHIPS Act politicians and the world, TSMC Arizona will NOT be a leading edge fab.
One thing lost in the shuffle with the TSMC Arizona fab news is that Taiwan passed a law recently requiring all leading edge semiconductor fabricaton performed by any Taiwan company (basically just TSMC) be done only in Taiwan. So, despite what TSMC told the CHIPS Act politicians and the world, TSMC Arizona will NOT be a leading edge fab.
One thing lost in the shuffle with the TSMC Arizona fab news is that Taiwan passed a law recently requiring all leading edge semiconductor fabricaton performed by any Taiwan company (basically just TSMC) be done only in Taiwan. So, despite what TSMC told the CHIPS Act politicians and the world, TSMC Arizona will NOT be a leading edge fab.
The Republic of China (Taiwan) has had the National Security Act since 1980s, which has been amended multiple times over the years. To implement this Act in practice, enforcement rules and regulations are established by the executive branch. This approach is quite similar to the U.S. system, where Congress passes Acts, and the executive branch establishes regulations for their implementation.
Under the regulations associated with the National Security Act, there is a list of "Critical Core Technologies" for which individuals, companies, and institutions must obtain prior approval before exporting or transferring them to foreign countries. The recent proposed amendment to the regulation (currently in the public comment period, if I’m not mistaken) aims to add 10 new categories to the existing 22. Many items on this list relate to military or dual-use technologies, such as satellites, biotechnology, material sciences, semiconductors, and computing. Cutting-edge semiconductor technology, defined as smaller than 14nm, is not part of the newly proposed additions since it has already been included in the existing list for some time.
TSMC's Phoenix, Arizona fab will begin high volume production within the next 1–3 months using N5/N4 processes (commonly referred to as 5nm/4nm technology). Clearly, these two nodes are more advanced than the 14nm threshold specified in both the current and new editions of the regulations.
It seems the Taiwanese government may be signaling: "Let's talk..."
This sounds standard for large mega projects of which I have worked on more than a few. Tackling problems on very large multibillion dollar projects was part of my job. It's surprising when a mega project doesn't run into a few problems along the way. A billion here a billion there is how these mega projects go. Major trouble shooting was part of my job. I hold three contractor's licenses in California.
1. They're so complex that no planning team can account for every single detail and a few (many) things get missed.
2. The "risk reserve" that is supposed to cover these kinds of oversights are usually reduced or even removed entirely for the "bidding" process (externally or internally) -- because of competitive reasons (i.e. make the sale).
3. Laws, regulations, and requirements can change over the years -- meaning you have different requirements when you build vs. when you start thinking about it. Legal complexity typically only increases over time..
4. Even a single delay in the critical path can have a multiplier effect in costs. (1 week delay in one component can mean 6 weeks of additional labor elsewhere because of inter-dependancies)
5. The team that implements these megaprojects is often different than the team that put together the propsosal or "bid". Inevitably, something gets lost in transition to execution, and the usual human error of "reinterpreting requirements" can add cost and schedule challenges. One could write a book on how requirements are handled being the root cause of these overruns..
I could write a lot more, but these are the typical things I've seen on $100M+ and $B projects in IT and Defense that transition and run above initial projections..
