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!
I wanted to clear this up for the forum. Above is a picture of an old processor core, an AMD Ryzen 7 3700X. Some things to notice about the markings on the outside of the package:
Diffused in USA (Global Foundries chip in the package)
Diffused in Taiwan (TSMC chip in the package)
Made in...
3.66 TW to 1 US fab ratio is what he's saying I think, as a gentle response to Chinese (BS) and Taiwanese (not-exactly-BS) concerns.
When you build this many fabs and make improvements with each fab, as TSMC does, it creates a wide moat with competitors. It's secret sauce, but not that...
11 fabs in 1 year! 1 per month, with February for Lunar New Year.
I'll make some predictions:
These will be the most sophisticated fabs in the world, built at the lowest cost, with the best yields.
No business has built this many fabs in such a short time before.
4x is probably about right, based on taking longer (about 2x) and other costs higher (about 2x). Over the course of a decade, if fabs continue to be built in the USA, that gap will reduce until it reaches parity.
I think we'll see Applied Materials shifting production of most tools from...
What the 100B investment means, some takeaways:
1-TSMC is satisfied with the ramp in AZ. This was always somewhat of a question, doubted by Morris Chang, and he was wrong.
2-TSMC AZ has room for 10 fabs. With this latest tranche, 5 have been announced. That's normal pace for a TSMC site in...
TSMC investing 100B in the US, it's a lot.
US currently has 12% market share:
It will take sustained fab construction for a decade to increase that 12%. Other countries are expanding. We need to expand quickly just to maintain that share. Right now, it is ridiculously hard to build a fab...
There are some issues with the national labs, tracing back to the publicly listed operators of said labs (SAIC for one). As with other defense contractors, they know how to get paid, get good results for SAIC, while the mission is less important.
National labs do attract and retain good...
"First, a world-leading Western chipmaker has to have its research arm located in the U.S. or the West, not Taiwan."
Portland advanced logic R&D could be converted to a national lab, like Sandia, Los Alamos or Fermi. The goal would be like Sematech, develop and implement a national...
My opinion is, it will take 10-15 years to build a trade as a Foundry. Intel probably doesn't have that long. I think the real question is, will there be a #2? Or does everything move over to TSMC in the next few years, Intel and Samsung Foundry virtually disappear, and then China becomes #1...
It's a good analysis of a good strategy which will probably succeed. It's not a coincidence, many leaders like Peng Bai, very experienced, know the scale at all costs, grab market share, then monetize later model well, because it's a Silicon Valley's model. The context seems foreign because...
There is some good news, the second generation discrete graphics cards are getting rave reviews. But it's probably still true that this is too little, too late.
The US has a preference for capital-lite businesses. That's the route AMD took and it was successful. But a second AMD would just...
"The key to all this is people"--Yes.
"Human glue"--the difference between taking the same tool from ASML and running very highly similar processes on it, and getting very different results, in Portland, Hsinchu and Pyeongtaek.
Well, maybe just in Hsinchu and Pyeongtaek soon.
The US should be able to beat China at SiC. The main inputs are pressure and temperature and time. A CVD process. With low enough energy prices, the US will win, but scaling up of production to reduce cost in the US is urgent now. And as the article mentions, 8 inch is the US standard while...
Bloomberg Businessweek had this article in Apple News recently: https://www.bloomberg.com/news/articles/2025-02-25/beijing-is-warming-up-to-the-idea-of-market-competition'
Making the argument that China has the stomach for creative destruction (Schumpeterian competition), while the US doesn't...
This is a new thought on semiconductors, I think: It's earnings quality (ie sustainable economic moats) that matters most.
"Many of these companies are large-cap stocks that once dominated their industries. Semiconductor chipmaker Intel INTC, for example, previously dominated the technology...
15 Stocks That Have Destroyed the Most Shareholder Value Over the Past Decade
These stocks have seen their market caps shrink despite a generally bullish market environment.
Amy C. Arnott, CFAFeb 25, 2025
Last week, I highlighted 15 stocks that have created the most shareholder value over the...
Ever since Apple switched to TSMC, the pattern of practically all major chipmakers using TSMC was set. It's an interesting dynamic. Entirely based on speculation, I think; you place a bet 5 years in advance, and hope it pans out. So far no one has lost money betting on TSMC, and the bets on...
I wanted to second something Daniel has been saying, when it comes to discussion about US manufacturing, it needs to come with US R&D. The US has steadily been losing logic semiconductor R&D for decades. This has corresponded to losing manufacturing. I can't say which one is more important...
TSMC should look hard at any deals to expand manufacturing rapidly in the US.
TSMC according to Morningstar has a price to fair value of 0.75. This is a major, major problem, as bad as what got Gelsinger fired.
The market discounts TSMC stock heavily over geopolitical risks.
I think P/FV...
Bunch of meanies :)
Exynos 2600 and Exynos 2500 die size is not available.
Exynos 2400 die size is 137.4 square millimeters (mm2)
Snapdragon 8 Elite (Galaxy S25 chip) is 169.8 mm2 on N4
SF2 is smaller than N4, so the die size will likely be smaller, but if it has AMD graphics, it might be...
