TSMC Earnings

TSMC Earnings

The Six Five team discusses TSMC earnings.

If you are interested in watching the full episode you can check it out here.

Disclaimer: The Six Five Webcast is for information and entertainment purposes only. Over the course of this webcast, we may talk about companies that are publicly traded and we may even reference that fact and their equity share price, but please do not take anything that we say as a recommendation about what you should do with your investment dollars. We are not investment advisors and we ask that you do not treat us as such.


Patrick Moorhead: TSMC earnings, bellwether of foundries. And in many cases, the chip industry. Dan, did they crush it or what?

Daniel Newman: It sounded so sarcasm, sarcastic though. Look, there’s two sides of the story. Did they beat? Yes, they did. Did they crush it? It depends what you’re comparing it to. Because if it’s to a year ago, no, they didn’t. So yes, on the top line, they outperformed. And then of course, you always got to do the translation between currencies, Taiwanese currency and US. But they did beat it on the top and the bottom, but their revenue is actually down one and a half percent on a year-over-year basis. But sequentially, revenue is going in the right direction. Why is this important? Well, the semiconductor industry as a whole, has been in a pretty long tail slide in most areas that aren’t AI. And of course, TSMC has been propped up a bit. They are early often an important partner in many of the AI architectures, especially as you move towards five, and they have a very strong ramp to three. And it seems that they’re even ahead on interest around two, which is some of what was exposed in the earnings call.

The company basically, did point out the global complexity. But a couple of interesting things, Pat, is one, they saw their HPC and smartphone now generated close to 90% of their revenue. So just those two categories. In the last quarter, this supports a lot of what we’re hearing, Pat, but IOT dropped almost half in quarter over quarter. So huge fall, as we saw the consumer IOT business drop in a really big way. Here’s another really interesting part, Pat, is 72% of their revenue is North America and only 11% is China. So that’s a surprisingly large distribution of revenue coming from the US, and that shows just the strength and dependency that the company has for US-based chip design or fabless design companies, Pat. The 3 nanometer ramp is happening faster than five from a revenue standpoint and that is even despite pricing discrepancies.

So overall, Pat, what I am taking away from this is that we have likely and kind of following the CEOs that we’ve talked to, that bottoming out of the inventory has happened. There’s an optimism going forward into the future quarters about strength, now that inventory supplies are going to need to be refilled. Of course, TSMC sees this early because they get order flow ahead. They’re one of the earliest as demand starts to pop up. And so that restocking demand is expected to pick up. But being very, very clear, AI again, seems to be the biggest beneficiary. They have a strong ramp on their 3 nanometer node and they’re strongly positioned to continue to support AI growth. And with all these things happening at the same time, it does look fairly optimistic for TSMC. The only thing I would say is the company did announce or has announced that its 2 nanometer mass production is expected in 2025. So we’re only a year out now from seeing three go to two, and from that being entering its mass production phase.

So a couple of predictions I have. One is smartphones and mobile devices. PCs are going to pick up in a big way in this year. The AI PC in the second half. I think I know a guy that did a big presentation at CES and said, “There’s a super cycle coming in the second half.” I think TSMC agrees and so do I. It doesn’t mean we’re not still going to be facing another challenging macroeconomic year, but silicon semiconductors are going to eat the world. They’ll continue to eat the world. Companies will find efficiency and productivity gains with the help of AI. That means more investment in AI. And long and short, Sam Altman is going to build all the factories anyway. So TSMC is in big trouble when that starts in 2026. That was a joke. Maybe, I don’t know. I can’t tell for sure. But it is kind of fun to think that Sam’s going to get into the business of chip manufacturing.

Patrick Moorhead: Oh my God, I forgot to put that on here.

Daniel Newman: I have to think though. I know. I squeezed it in so you could comment on it. I have to think though that when he says, to build a network of fabs, maybe he meant to create an open relationship with some different manufacturers and create a network of fabs and use Arm and Synopsys, and maybe Intel and these different companies, to help him build capacity with some level of commitment that might be required on his part. Do you think he’s going to really get into the full IDM or is he more thinking fabless chip design with a network of fabs around the world? I don’t know. It was hard to read the way it was presented, but it was really fun to think Sam Altman’s going to be spending his next few years flying around the world building semiconductor manufacturing facilities.

Patrick Moorhead: So I think three things could make sense, and I air this out on X this weekend. What could make sense is for him to pull dollars together to build an ASIC that anybody who runs an open AI process would be able to take advantage of this and it’d be the most efficient. So you get the ASIC advantage over the GPU, and then you get an ASIC that is built for a GPT6 or something like this, now you’re talking. And maybe they could give some price reductions to those people who might want to go in with them on it. Or OpenAI or some affiliated business ends up selling these and the investors could take part of the profit.

The other thing that could make sense along with that ASIC is to do what Apple does, which is to take CapEx to either get bleeding edge or leading edge. And I think if you applied some bleeding edge CapEx for the training, and then potentially, some leading edge for the inference chip, let’s just say there are two different chips, yeah, you might have something. I mean, listen, if Sam was going to create an IDM, he’d first have to be a designer, which he’s not. And secondly, he’d have to have access to high EUV, which would be pretty darn expensive. So again, I don’t see too much there, but we’ll see. I mean, Sam Altman’s kind of like an Elon Musk guy without the science, but a business guy. And if anybody could figure out something that’s interesting, it’s him. Dan, you’re quite the biz dev guy. You should give him a jingle and ask him about that.

Daniel Newman: I’ll call him. Yeah, no.

Patrick Moorhead: No. Okay. By the way, back to TSMC, I mean, the thing that really blew me away, 50% of their revenue is leading and bleeding. By the way, I have to put asterisk. There’s no 3 nanometer transistor width here. 3 nanometers is 15% and 5 nanometer was 35% of the revenue. It does annoy the heck out of me that a high performance computing or HPC is PC, plus data center. Come on, man, break that apart here folks. Make it a little easier for people to know what’s going on. But all in all, looks pretty good. I want to add one final thing. TSMC did delay its $40 billion chip project in Arizona for another year, saying that it is waiting for US grants. I’ll say that’s probably a half-truth. They are behind and they don’t know how to actually get American workers to come in and do what they need to do, even though companies like Intel, and GlobalFoundries, and Micron have figured that out.

Author Information

Daniel is the CEO of The Futurum Group. Living his life at the intersection of people and technology, Daniel works with the world’s largest technology brands exploring Digital Transformation and how it is influencing the enterprise.

From the leading edge of AI to global technology policy, Daniel makes the connections between business, people and tech that are required for companies to benefit most from their technology investments. Daniel is a top 5 globally ranked industry analyst and his ideas are regularly cited or shared in television appearances by CNBC, Bloomberg, Wall Street Journal and hundreds of other sites around the world.

A 7x Best-Selling Author including his most recent book “Human/Machine.” Daniel is also a Forbes and MarketWatch (Dow Jones) contributor.

An MBA and Former Graduate Adjunct Faculty, Daniel is an Austin Texas transplant after 40 years in Chicago. His speaking takes him around the world each year as he shares his vision of the role technology will play in our future.


Latest Insights:

On this episode of The Six Five Webcast, hosts Patrick Moorhead and Daniel Newman discuss Apple Vision Pro developers losing interest, U.S. awards Samsung and Micron over $6B in CHIPS Act funding, does AMD have a datacenter AI GPU problem, Adobe’s use of Midjourney, Samsung knocks Apple off of number 1 market share, and Arm says CPUs can save 15% of total datacenter power.
In Recent Years, the Concept of a Sovereign Cloud Has Gained Significant Traction Among Nations Seeking Greater Autonomy and Security in Their Digital Infrastructures
The Futurum Group’s Steven Dickens observes that Oracle's recent $8 billion investment in Japan not only expands its cloud infrastructure but also strategically aligns with the growing global trend toward sovereign cloud solutions.
Hammerspace, Seagate, Quantum, LucidLink, and Resilio Are Among the NAB Products of the Year for 2024
Camberley Bates, VP of Data Infrastructure at The Futurum Group, covers the significance of data infrastructure at the NAB Show 2024 and the Product of the Year Awards.
The Circular Economy Can Help Your Company and Business Ecosystem Improve Sustainability Efforts and Create New Innovation Opportunities
Olivier Blanchard, Research Director at The Futurum Group, encourages you to download the latest whitepaper from The Futurum Group, which highlights circularity—a perfect read for Earth Day.