Certify Alexametrics gif

    Semiconductor Industry Strategic Trends

      Sep 29, 2021

    VLSIinsiders' Cloudside Chat — September 29, 2021

    This is the talk Dan Hutcheson gave for the Semiconductor Industry Association’s state of the U.S. semiconductor industry webinar. In it, he discusses…

    • How the Chip Innovation Engine Relentlessly Drives Opportunity
    • Semiconductor Technology Opportunities
    • Current status of the chip shortage
    • Long-term impact of US semiconductor policy
    • Why wafer fabs are technology treadmills
    • How we got here and the companies we lost
    • Why Investors are voting against US Chip Co’s
    • Policies to strengthen US Chip Manufacturing

    About VLSIinsiders: Every week our analysts have a cloudside* chat to discuss current events and key issues of concern while sharing what they’ve heard over the past week from the semiconductor industry insiders.

    * cloudside chat: A fireside chat without the fire and is across the clouds of the internet

    Other VLSIinsiders


    VLSIinsiders' Cloudside Chat — September 29, 2021

    DAN HUTCHESON: Good morning. I'm here recording the talk I'm about to give to the Semiconductor Industry Associations as data of the semiconductor industry webinar that they give every year. I talked last year. And so, I'm following up this year and then talking about. This year, I'm focused more on what's happening with industry policy and how the government sometimes is helping it and other times it's restricting it. So, I'm going to talk a little bit about that and the history of how we got here. So let me share some slides with you. So, if you look at this beautiful shot driving into Death Valley, this is my way of depicting what's happened. We have great opportunities, but sometimes they really are handcuffed by policy. And when I say great opportunities, if you look at what's going on in the semiconductor industry in terms of the various technologies that are being developed, and how they've changed things at the macroeconomic level, how chips have gone into everything, everywhere. They're affecting everything from being able to find vaccines for COVID to national security, and then down into the layers below that into the equipment and materials, and even into the sub systems that are made to go into the equipment. We'll spend time here, I could spend all day here. And then if we look at the technology opportunities in terms of what's being developed by our industry a lot, they cover the DRAM and NAND, and then the logic, analog, and power or LAP, if you will, in terms of short acronym. When it comes to this year, this year was really difficult, because, you know, last year, everyone thought the industry would be coupled to the macro economy going into a downturn, but it didn't. Things stayed pretty much tight all year long. And then in the first quarter of 2021, we got into this massive shortage and it really affected the auto industry and it still is in shortage conditions even though it's beginning to work its way out down here on a weekly basis. And if you come back, the only area where we've had over supply is in DRAM. So that really was just sort of a mistake because everyone cut back production in early 2020 and then that came to haunt us later on.

    If we then look at the long-term impact and talk about, you know, some of the national issues about what's happened when the United States has lost manufacturing share for six decades, it goes all the way back. We, our data goes as far back as the 60s, when the US have actually over 90% of the world market. Europe and Japan were on the rise because the US Justice Department had insisted on allowing the patents, Bell Labs patents, to be patented, to be given away for free to other countries. And so, you see that go and then Japan is the one that really rises. Europe never really got above 20% or big peak was in the late 60s and then things started to decline as Japan rose because Japan was where everything was being made. And there was a shift of manufacturing out of the West out of Europe and America and to Japan. And so that drove them up until around 1990, and what you see here is this sort of recovery in United States which was driven by Sematech. But anyway, we get to this point here and everybody starts going on down. But we have the rise of Taiwan and South Korea, and more recently, the rise of China, which now is manufacturing about as much as the US has manufacturing. So, this is an area of wafers, you know, we can get into other ways of measuring it in such as what's at the leading edge or whatnot, but this is area wafers. So, if you look at what that why that happened, to a great degree, it happened because of the law of Moore's law on demand. And because really, wafer fabs are natural monopolies. You can use technology treadmills, which is Cochrane serum goes back on agriculture. And essentially, you always have to invest in these new fabs. Today, it cost about $20 billion for a new three nanometre fab. And you know, if you have a single delay in construction, that's an $11 million depreciation. It's really crazy what you can lose very quickly. And then once you've got the new fab, if you want to upgrade to the next node, it's a lot cheaper to upgrade to the next node once you're there is not providing you can pull the research and development off to make it happen, so that's why you see increasing concentration. There's been other reasons as well. But if you go back and look, you know, this is my, one of the favorite charts that I give out that called the Pac Man chart and essentially what happened is we go from 2001, 130 nanometers to today. We go from, you know, lots of companies in logic and SOC and memory, dwindling down to about four memory suppliers and only five. or excuse me, only three logic manufacturers at the leading edge. And here, I don't include some of the others. This is just leading edge, it's not companies that are playing catch up. So, there's a lot, there are other companies, but they are not playing at the leading edge. And what caused this to happen was, first of all, we had EDA allowed to design for coupling to be broken. So that meant that you didn't have to have a fab to make semiconductors, and then the silicon shield so that the incentives that the government put in place and South Korea and Taiwan really drove those areas to have a fundamental competitive advantage. And of course, the United States, you know, the investors there was this capex phobia that really favor fabless companies. And you could push the capex off to a third party that drove the whole fabless phenomenon, the foundry phenomenon. And then there was also a phobia in the United States government against industrial policy and that favored all of the regions that were practicing the silicon shield. Now, the silicon shield was simply a way for those countries to protect themselves from the recognition of China as an independent nation by the United States. So, we didn't help ourselves there. And then of course, there was all the foreign STEM students who have come here for years. They've had a financial advantage at US universities because they pay more than local students do, so that's allowed them to gain more knowledge base and more grow their whole human capital at the expense of US manufacturers.

    Now, and then the more recent one is the idea that we can raise corporate taxes in the last year. If you look at the impact of US semiconductor companies that VLSI tracks versus the non-US companies in our indices, and our stock indices. Last week, it was a 13-week, 13 percentage point gap, a week before it was 15 points. And the only thing that happened was is when GE came in and started, you know, they shut down cryptocurrencies, they shut down some of these other things. And so that whole strategy of raising taxes, though, has cause investors to look elsewhere to put their money because they know that the earnings are going to go down. So finally, if we want to talk about policies that we can do in the, at a governmental level, to strengthen US chip manufacturing. Number one, besides it's in the plan. Now, you know, I know that there's a lot of money coming in the industry, but number one, which is simply have a single federal construction permitting process. Because wherever you go to build a fab, those differences in permitting and the differences in the knowledge base of the local permitting, authorities can really slow fabs up and could cost a lot of money and put you at a competitive advantage. This is an area where Arizona has really taken a lead. The other one is to continue to bolster the public private partnerships like what we see in Albany, as well as with the SRC, the and then, you know, we did have Sematech at one point. Europe's done a marvellous job with IMac. Then, you know, the other one I would say is to link our leadership and our universities and semiconductor technologies, the SAA shown time and again, the US still leads in technology, it's just not leading manufacturing. So how do we have that manufacturing size? We need to have those that link up with universities. And then finally, keep corporate taxes low, which I just mentioned. Now, the other side is when we talk about linking the universities, it's how do we drive STEM? You know, science, technology, engineering, and math, and how do we drive those fields so that we get a human capital based in the United States that can support manufacturing. You know, and I think we need to provide some form of financial incentives for staff. For example, one thing that I think we could do right away is just simply allow those college loan payments to be tax deductible for people that got a STEM degree, especially if they're working in the semiconductor industry. And then finally, foreign students who are getting an advanced degree in STEM, why don't we just automatically give him a green card if they want to work in the semiconductor industry and if they want to stay here and then put them on a path to citizenship? You know, so many of the great leaders of the American semiconductor industry came from these foreign countries, like Andy Grove or Dan Madden. They came from behind the Iron Curtain, got educated here and then went on to lead some of America's great companies such as Intel and Applied Materials. So, I want to thank you for taking the time today and really appreciate and appreciate you coming to our site to see this. Please look at our app where you can get VLSI Chip Chirps, which are just analyst’s items on what's going on in the industry as well as what some of our videos you can also download them to your phone. So anyway, have a good rest of your day and thank you very much.