The US Government is set to spend billions on American semiconductor manufacturing. Is it an aggressive, much needed, offensive against aggressive, international dominance or destined for disaster? Few industry leaders know where this is headed like Ray Zinn. In this Tough Things First podcast, Ray explains where the industry stands and what it really needs. A do not miss podcast for the industry and government leaders. (Watch the full video podcast HERE)
Rob Artigo: Welcome to The Tough Things First Podcast, your indispensable source for business leadership and life advice, with the longest serving CEO in Silicon Valley. I’m your guest host, Rob Artigo, and he’s Ray Zinn. Hello, Ray.
Ray Zinn: Hello there, Rob. So good to be with you this morning.
Rob Artigo: And, it’s great to be back as always. So, this is obviously a special edition of The Tough Things First Podcast. If you’re listening right now, and not watching it and you’d like to, you can go to toughthingsfirst.com, find this podcast there. You might even be listening to it there, but there’s a link. You can click the link. It’ll take you to Ray’s YouTube channel and you can watch, as well as listen, and let us know what you think.
Well, Ray, the White House says, “The Chips and Science Act of 2022 will be a historic investment that will poise US workers, communities, and businesses to win the race for the 21st century.” $52 billion for the Chips and Science Act that the Biden administration says, “Will boost American semiconductor research, development and production, ensuring US leadership in technology that forms the foundation of everything,” as we’ve discussed many times on this program.
Ray Zinn: Yeah.
Rob Artigo: So, is that what’s going to happen?
Ray Zinn: Well, let’s do a little background first. Let’s kind of drop back and take a look at the industry from a 40,000 foot perspective.
Rob Artigo: Okay.
Ray Zinn: So the industry began in the ’50s. The semiconductor industry that is, began in the 1950s. Then most of the companies were in Silicon Valley or the Bay Area. And there weren’t too many of them. There were maybe seven or eight of them at the time. Not even that many, maybe four, actually. So then they started out making just transistors, which is basically a three liter device with a base collector and emitter, and not much of a technology except was replacing the vacuum tube.
In fact, Varian was making vacuum tubes in the Bay Area at the time, and that was for the radios, and TVs, and so forth. They were vacuum tubes. But then the semiconductor industry came along and managed to make that capability really unique, were making them very, very small. So with the advent of the development of the transistor or the semiconductor, we’re able to make these electronic devices extremely small, because the vacuum tubes were huge. They were anywhere from two to three, to five inches in size, and doing a single function as you would.
So anyway, then in the 1960s, early ’60s, we developed what we call the integrated circuit, and it was nothing more than a flip flop. Meaning it was a transistor, it would flip back and forth. And so, that was the advent of more than a single function on a chip. And then from the development of the flip flop or the first IC in an integrated circuit, in about late 1960s, Bob Noise, who was running Fairchild at the time, left and formed Intel Corporation. So a lot of companies spun off in that late 1960s timeframe.
But in the seventies, with the advancement of semiconductor technology, the Japanese came in the picture. The Japanese have always been well known for electronics. So when we developed the first computer, which was back in the mid-seventies or early eighties, in fact, I think the first PC was developed in 1981, it required memory. Lots of memory. Now, memory has a lot of repetitive functions, and in other words, the chip is jammed with just these little flip flops, millions of them. And so, defects became a real issue. And so, the yields were not too good, meaning that the production rate, or the number of good devices on a wafer were not that great.
And the Japanese were able to, because of their standards they had, their more disciplined, cleanliness issues. They were able to produce these memory chips much cheaper. While the labor market was cheaper, there were less environmental constraints in Japan. And so, the Japanese really took over the memory market. And because back in the early eighties, memory was very, very expensive. In fact, that’s why Intel developed what they call the, oh, my. Just went brain dead for a minute. But a CISC, which is a complex instruction set, and that the reason is because they wanted to reduce the cost of memory. So Intel developed with what they call a more integrated instruction set, so that they didn’t require lots and lots of memory. And that was somewhat to combat the cost of memory. Memory was very, very expensive. And only a very few companies in the US actually produce memory. Most of the memory was produced in Japan.
So in the middle eighties, the government, along with the industry, Semiconductor industry, put together what they call SEMATECH, S-E-M-A-T-E-C-H. So SEMATECH, again was funded kind of 50/50 between the US and the large semiconductor companies, to develop advanced technology and to increase yield and reduce costs. And so, the government pumped in a hundred million dollars a year into SEMATECH. And the industry corresponded with their roughly a hundred million dollars. So $200 million a year into this joint venture with DARPA, Defense Advanced something or research department to develop advanced semiconductor technologies, and to definitely improve yield and reduce cost. And so, this is not the first foray by the government into trying to help the industry compete.
And so then, the industry began to proliferate. And because of the strict environmental constraints in the Bay Area, Silicon Valley, San Jose, Sunnyvale area, California, the companies began to move out. In other words, semiconductor companies began moving out of the Bay Area because of the cost of just the environmental constraints. That goes back to this trying to clean up as you would, the water and the air and so forth. And that forced a lot of the companies to leave. And today, in the semiconductor in Silicon Valley, there are very, very few semiconductor companies, maybe just a couple left, that actually produce semiconductors from wafers, anyway, in the Bay Area. They’ve left. They’ve gone to Colorado, they’ve gone to Texas, and New York and other places where it was less expensive to produce semiconductors. And in addition, a lot of proliferation was going on in the Far East, not just in Japan, but in Korea, Singapore, in Russia, China. Mainly China, I would say. India.
And of course, Europe also had a semiconductor industry, but they had a lot of environmental constraints because the Europeans were very environmental focused, both clean water, and clean air and whatever. But the industry introduces a lot of toxicity because of the nature of the production of semiconductors. And so therefore, we have this great move out, as you would, of semiconductor companies from Silicon Valley. Silicon Valley no longer is really Silicon Valley, it’s more Software Valley, because again, all of our environmental constraints, software doesn’t require any toxicity or any manufacturing. It’s just people developing software. So, Silicon Valley is now really Software Valley.
But anyway, back to our original pursuit of what the Biden administration’s doing with this 52 billion. Again, looks like we’re reinventing the wheel as you would, because now we’re going to go back to doing this joint venture with the industry and trying to develop more competitive technology.
But we should have learned from what happened in the early eighties and through the nineties, because the semiconductor industry is more individual capitalistic companies. They don’t want to give away their secrets, and they want to maintain that secret sauce, in both in manufacturing as well as in design. And so, they don’t really work very well together. Competitors honestly never work well together. They don’t. Well, when we talk about compete, compete means you don’t talk to each other. As you know, in the government right now we’re talking about top secret stuff that may have or may have not gotten out. And so, that’s the same thing in companies is they like to keep their top secrets, top secret. Apple is very much that way. I mean, they can’t learn anything from Apple. Apple wouldn’t work with anybody. And so, these companies didn’t really work very well together as far as trying to make and develop better semiconductor advanced technologies.
Rob Artigo: That was it though. Excuse me, with SEMATECH. One of the major hangups was the fact that they wanted people to do their, “Okay, you have your R&D team and you have to share that information with this R&D team.” And I would think there would be two. There’s one, the biggest hurdle is what you just mentioned, which is, “Why would I want to invent something and give it to you? If I’m paying all the money, I’m doing all the R&D on this, it’s costing me all this money.” And the other part of course, is that if you have the smartest minds in the area that’s in design of the chip, in one room sharing information, it seems like if you’ve got 10 companies and those 10 teams are together in one place, sharing that information, it naturally will migrate towards something. And you’ll have kind of a group think, instead of the many different tentacles that come up with these brilliant ideas that have advanced semiconductor for so many years. You had those individual pods where people created and masterminded their own products. And that’s how they came up with something new.
Ray Zinn: Well, Micrel, the company that I formed in 1978, we decided rather than try to make our designs unique, we tried to make our process unique. And so obviously, we didn’t want to share that technology outside of our company. And so even though some of the technology did get outside of the company, when people left and went to another company, we tried to keep it as closely held as we could, to give us a competitive edge. That’s the problem with SEMATECH. These companies really didn’t work all that well together. They did some things which were joint venture, for example, process technologies. And this equipment that that’s needed to manufacture these advanced processes is very, very expensive to develop. And the manufacturers, the people who make this tech, semiconductor equipment, they need assistance from the semiconductor industry to make sure that they’re developing what they want, what they need. And so, SEMATECH did help out in that respect, because they were able to work with the equipment manufacturers to develop advanced processes. And so, that did help and that did advance the industry.
Rob Artigo: And again, Ray, that was again, a matter of competition. And you had Japan, that was really the target, because it had taken over everything and the US wanted to be more competitive. Which now, the government wants to have the industry come back and be more competitive to face their technology foes, which includes, like you said, basically all of Asia. Back then, they spent a billion dollars in government money. It wasn’t a complete failure. Like you said, SEMATECH did something, did add something to it, but a billion dollars in today’s money was spent, basically, to do that. And we’re looking at infusing $52 billion now. Are we going to run into the same situation, where it’s going to fall short, where they had hoped that it’ll be great?
Ray Zinn: Well, that’s where we want to end up is, is talking about this new idea, the 52 billion. I’m not sure how they’re going to disseminate that money, without causing a problem amongst these companies. Because there’s probably five or six semiconductor companies that have their own wafer fabrication manufacturing, that would really benefit from this 52 billion. I know maybe seven, at the most, in the entire world. I mean, US companies, I’m talking about throughout the world. And so, the rest of them are fabless, meaning they don’t have wafer fabrication manufacturing, and they use companies in China, and Korea, and Taiwan, and Vietnam, and also in Europe. But, there’s probably over a thousand semiconductor companies that are fabless, whereas there’s only seven, in the US anyway, that actually have manufacturing capability. So, are you just going to help out the seven that have manufacturing capability?
One of the things we want to do is get the technology back into the US, get manufacturing back here. That’s the challenge is, how do you do that? So my thinking is, don’t know if this is going to be seen or heard by the political leaders, but my thinking is that if we’re going to use that 52 billion, it to be encourage all semiconductor companies, whether they’re fabless or fabbed, as you would, to manufacture their chips here in the US. There’s how that money should be spent. So it may be, it’s an incentive. Maybe you allocate it based on revenue dollars or maybe based on the percentage of your chips that are made here in the US, that you would get then a percentage. For example, if you do 85% of your chips here in the US, then you would get 85% of your allocation. If you only did 10% of your chips here, then you only get 10% of that allocation, to encourage them to actually manufacture back here.
And I would do it through tax incentives, not just a handout. Handouts never work, as we know. That’s socialism. So handouts don’t work. So handout, I’m not in favor of. Tax advantage, yes, that’s the way I think it should be done. Allocate that money, 52 billion in terms of tax incentive. So that would encourage the semiconductor companies, especially the large ones, to come back into manufacturing in the US. That’s the whole purpose of this Biden semiconductor 52 billion deal. So there’s probably six or seven semiconductor companies in the US that produce over half of the semiconductor dollars in the US. So just a handful. As opposed to that, there’s probably five or 600 fabless semiconductor companies that do not use, very few would use US manufacturing. They use offshore manufacturing, either in China, or Korea, or Taiwan to manufacture their chips. But we need to get that technology and that manufacturing back in the US, and I think it should be tax incentives, not handouts, to get these companies to want to manufacture their products back in the US.
And finally, before we close off this podcast, we need to reduce the cost of manufacturing here in the US, through all these environmental constraints. I know we’re having that with the fuel industry, the petroleum industry is having a lot of constraints put on them. Same thing with people manufacturing good food, packaging companies. We have a lot of constraints. We have a lot of environmental constraints that just make it difficult for our industries to manufacture here in the US. So we need to kind of slim back that. I mean, environmental issues are important. No mistake about it. But if we want to continue to be competitive and to have top notch electronic products, we have to be somewhat lenient, more lenient and less restrictive on some of these environmental issues to make it cost less to manufacture here in the US. So we need to incentivize these companies to come back, to do more manufacturing here, whether you’re fabless or fab.
Rob Artigo: All right. Last question and just a really brief answer on this, does that $52 billion get spent eventually, and do you think it ends up being a good investment for the US government, or US taxpayer?
Ray Zinn: Well, I don’t think it’s going to get spent correctly. I mean if it does get spent, it’s going to be just a waste like Solyndra. It’s just going to be a wasted government expenditure. As I’ve mentioned, the way we ought to do it is through tax incentive. Encourage these companies to come back to the US by giving them a tax incentive. And if it is, we’re going to put all these environmental issues on them, then the government ought to give them some kind of a tax reduction for manufacturing under tough environmental. We can do that. I mean, it’s certainly a matter of choice in our part, to encourage these semiconductor companies to come back to the US, manufacture. And if we’re going to have these tougher environmental issues to deal with, then the government helps them. And so, we can be competitive. That’s the whole game. The game is competition, being competitive. That’s how you get us back on top again.
Rob Artigo: Well, I hope the White House does hear this and give your thoughts a real … I don’t know if they’re going to do much discussing it there in the White House, but somebody out there in the government will have to decide how to expend this stuff, and maybe they’ll listen to you. I hope they will. Thank you, Ray.
Ray Zinn: What we ought to do, what those of you who are watching and listening to this program, please forward and share this podcast with your political leaders. Would you do that for us, please? Thank you.
Rob Artigo: Good advice there, Ray. As always, you can reach out to Ray Zinn with your questions at toughthingsfirst.com. You can continue your education and the conversation with all of the podcasts through blogs and links to information about Ray’s books, Tough Things First, and the Zen of Zinn, one, two, and three. It was a great conversation, Ray, and a great video podcast. So I hope people will rate the podcast, will tell us what they think about it, and we’ll come back and do another video soon. What do you think?
Ray Zinn: Well, number three is coming out. It should be out in next month or two.
Rob Artigo: All right. So, Zen of Zinn III.
Ray Zinn: Yeah, That in fact, Zen of Zinn III has been printed. Here’s the book.
Rob Artigo: All right.
Ray Zinn: Yeah. This is not an official copy. It’s a test copy, as you would review, but this is Zen of Zinn III. It is ready to be produced. We’re just waiting for it to get published.
Rob Artigo: Okay, awesome. Thank you, Ray. I really appreciate the conversation. I think I learned a lot about the industry, and what the government’s doing here with this program.
Ray Zinn: Thank you.
