The Six Five team discusses U.S. Awards Samsung & Micron over $6B in CHIPS Act.
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.
Transcript:
Daniel Newman: We had a couple more CHIP Act Awards this week.
Following the seat at TSM, it seems like the $6 billion is the range. Everybody’s going to get $6 billion. You get $6 billion, and you get $6 billion, and you get $6 billion. Oh, and Intel, since you’re the most important, you can have $8 billion. I don’t know Pat. Look, I think you and I have both been pretty steadfast about this correlation that exists between memory technology and AI chips, and that building the advanced. Mostly we’re talking about HBM right now, at scale is going to be a rate limiter for the success of AI silicon, because you need enough memory for all the throughput that you’re going to require to deliver on the promise of AI. So who’s building that?
Well in the U.S., it’s primarily Micron is the US-based leader. And of course Samsung is another major player in this space. So both companies are getting more than 6 billion. Samsung’s going to be expanding. It’s local here, Pat, presence both in Taylor and Austin. It’s largely going to focus on what people expect to be it’s four, and then later two, nanometer process and its most advanced process for memory.
Now, Pat, interestingly enough, all the news keeps banging on the fact that US-based manufacturing of Silicon has dropped from 37% to 12% over the last three decades. Interestingly enough though by the way, this is just a reminder, you and I both know this, but for everybody out there, the actual percentage of leading edge built here in the US is…
Patrick Moorhead: Very small.
Daniel Newman: It’s like zero. It’s negligible is basically the amount. And so even when we try, when Reuters tries to say it’s 12%, that’s mostly older nodes above seven. So, the fact is that what we’re trying to solve, for those that have forgotten because it took a couple of years from the time that the CHIPS Act was announced by the time that these grants started going out, we’re trying to solve the fact that we have no resiliency outside of Taiwan for anything seven nanometer or below. So all these grants are heavily focused on those processes and it’s obviously logic and memory.
So with TSMC, you’re seeing the focus on logic with Intel Logic, with now we’re starting to see the memory companies with Samsung and Micron getting substantial awards. And there are others, some of you probably heard of SK, they’re not, so far, have not been granted any dollars here. But for Micron, it’s largely expansion in the Idaho and New York markets. They have a big Four fab project in New York. So they probably got a lot of support from Chuck Schumer, Senator Chuck Schumer, for that.
We know that New York has an interesting semiconductor out in Albany. We’ve been to the IBM Research Center there. We spent some time, but they’re really trying to be another hub of innovation there. So, this is going to obviously expand it. We know that Global Foundries also has a pretty substantial footprint out in that part of the world.But Samsung, it’s all Texas, baby. It’s all here in Texas. This is where it’s all happening. This is where you should move right now, viewers. You should be down here in Austin, as long as you tend to agree with Pat and I’s viewpoints. But look, it’s moving forward. Awards are being granted. You’re talking about some different timelines with each of them.
And so it’s going to be several years, though, before we’re starting to see the byproducts of these. And by the way, both companies, I think at Samsung’s $40 billion is what they’re planning to spend all in. And Micron, I think it’s a hundred over two decades, though. So, good progress on the awards. I’m still wondering when at what point we’re going to start to see real, meaningful production here in the US It’s at least a few years out.
Patrick Moorhead: So memory market is back. And for those who have not been following for the last 40 years, memory is hit-and-miss. You’re either in an oversupply or an under supply. When you’re in oversupply, your customers are driving rapid price cuts and it’s ridiculous. And we saw negative gross margins for Hynix, Samsung, and Micron. But memory’s back, baby. Why? Because all of the platforms that it goes into are gaining demand, and all three of these companies pulled back on their capital expenditure, and even pulling back on the amount of hours that they run their foundry.
So prices are going up. You may have seen that Micron basically announced they were sold out of HBM3 for the year, and it was March when they made that announcement. We saw Samsung come in from a P&L standpoint on semiconductor. Now they don’t break out memory versus logic, or sorry, memory versus things like processors, but you just know that most of that was memory because most of the processors they manufacture, they consume themselves.
So it’s great to see this memory, this back. And oh by the way, wait until AI PC and AI smartphone hits in with all this on-device models. It’s going to be absolutely bonkers, because you need more memory to do the processing of it, and you need more storage to store the 10, 20, 30 models that are going to be sitting on your phone. And then from a Samsung standpoint, here in Taylor, it’s super, super exciting. And essentially what that means is that you’ll have Intel coming out with 18 a first here in the United States.
You’ll have a two nanometer coming out in Taylor by 2026. And then I think TSMC said by 2028, they would have two nanometer running in Arizona. I’m still piecing together if TSMC is being conservative on their dates. I had originally thought that to get the money they would’ve had to do bleeding edge here. And 2028 for Samsung, two nanometer is anything but a bleeding edge, it would be one or one and a half nodes…
Daniel Newman: It’d be like seven right now, right? It’d be like saying we’re at the leading with seven?
Patrick Moorhead: Probably more like five. Three nanometer is leading edge. And by the way, this nanometer stuff is fake. I am tired of doing air quotes around it, but if somebody can use an electron microscope and show me a two nanometer gate width, then I will retract. I will retract that.
Daniel Newman: 0.1.
Patrick Moorhead: Macro… Sorry, what’s that?
Daniel Newman: I was joking like 220, 221.
Patrick Moorhead: So macro here, US by 2030 has a… They want to do 20% of chips here in the United States, and that’s how this all comes together.
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.
