Reviving U.S. Semiconductor Supremacy: Education, Reshoring, and Ohio's Tech Renaissance

We have two really great speakers here today Dr John Berry , who is the president of Central Ohio Technical College . John , welcome Greetings , thank you all . And Ashwin Rangan , who is formerly Senior Vice President of Re-Engineering and Technology at Connexant and an old friend of the show . Ashwin , welcome back , thank you .

So , purely from an economics perspective , it takes the typical input factors of you need the right people , you need money and you need the equipment . When the act was passed , none of those three were in a state of preparedness to take complete advantage of the act , and the act was passed in 2022 . Here we are right in the middle of 2024 .

The money is only just starting to flow toward people who are interested in making their own contribution outside of the subsidies that they receive . So now there is talk about all right , assuming that the shell for a wafer fab is available . That's a big assumption because that says that people know how to build one . Where is the equipment going to come from ?

How fast can the equipment be made and how soon will it arrive here ? Esml equipment that is responsible for deep UV lithography , as it's called , costs about $350 million give or take and is the equivalent of about 10 trailer trucks . One piece of equipment for one step of manufacturing .

Well , and one of the things that actually was the trigger for this show was a comment that I heard you make recently , ashwin , that when the Intels of the world and the TSMCs of the world Taiwan Semiconductors of the world were considering the notion of building plants in the US , they went looking for talent to be able to actually do the construction work and

then ultimately do the operational work , and they couldn't find the people . So that kind of brings us to the question on your side , john is that , what is the need out there as you're seeing it in central Ohio , and extrapolate that out to the US ?

This is not something of which we had a rich history , a strong foundation within , but what we do have is a significant foundation of advanced manufacturing . That has been a strong point of Ohio , central Ohio especially . It's still been a very healthy element .

If you think about the state of Ohio and the whole notion of Rust Belt , with so much of the industry having moved out , this has always been a core and has maintained a strong level of significance .

Companies like Boeing , those sorts of elements that we've been doing production engineers to support that network Now extrapolate that into what we're talking about with semiconductors . We had to be able to first understand the industry . What is it that this production requires ?

And Ashwin hit it on the head this takes a lot of initial support to generate and to design and for us , our charge was to produce the workforce element , both the engineers who are operating the functional systems within the fabrication centers and then the construction of the sheer size and scope of this fabrication design we were talking a little bit ahead of time

about the sheer size of this construction project . It is the singular largest construction project in the history of the state of Ohio , so just providing enough workforce of skilled tradespeople , those who are in construction , those who are doing plumbing , electric work , that has been sort of a heavy lift , but it's going fairly well .

The foundation for this fabrication center the first two designs are well underway . They finished their basement , which was truly the foundation of that process , and now we are really trying to ramp up the production of the people to work the fabrication center .

You know , that's the antithesis of what intel kind of brings to the map in this world . At its core , this is advanced manufacturing . It is the manufacturing of these wafer chips in this process . But what the skill sets of individuals at the technician level , which is about 70% of the workforce that will work in the fabrication center ?

They're actually just advanced troubleshooters . They are looking at the enormity of this equipment that produces the chips themselves , which is mostly automated . But they have to be the instant response to any time any element goes down . If there is any shortage of the production line , they have to be the first responders .

So they have to be trained in that skill set of troubleshooting , knowing how to use the tools and networks to respond to that . So we have to produce individuals who understand engineering components at at least its base level and then work in strong teams and units to respond to any shortage of production .

It was a facility that came up somewhere in Idaho that we made that investment in and I think they're trying to expand that on the back of the CHIPS Act as well .

There was a point in history where Japan was regarded as a potential technological threat and multiple initiatives were undertaken by the United States , both from a government perspective and from a conglomeration of semiconductor wafer manufacturing companies at the time , to confront that particular threat that came from , that perceived threat that came from Japan .

Japan had already made a name for themselves in miniaturizing technologies and therefore seemed to have the native skill set to take anything , even wafer fabs , and perhaps do it better .

So on the back of that , organizations were created that were spanning across players and looking at how to advance the state of technology and advance the state of knowledge taken to make them manufacturable , those advanced technologies .

Then something happened roughly around the beginning of well , really in the middle of the 80s really in the middle of the 80s Morris Chang was working here in the United States a Taiwanese gentleman and he looked at the state of the wafer fabs and he decided that Taiwan might make for a good place to create a wafer fab , because Taiwan at the time was looking

to make investments in becoming a part of the supply chain . He went back from the United States to Taiwan and created what today is TSMC . Tsmc started as a contract manufacturer of silicon and that's all it was at the inception point .

However , as is very seen very often with such companies , they tend to climb up the value chain slowly but very surely , and today they're at a point where they define the manufacturing state of the art in terms of what is required from an equipment perspective to go to the next level of more detailed design , and that becomes an input parameter for the manufacture

of equipment from different parts of the world . So what we have seen on the United States side is that , you know , as recently as 35 years ago , we made more than 50% of chips here in the United States . Of course , chip demand was far lower at the time , but we practically made half of it .

Today that number is hovering around 10% and , after all the investments that we're talking about , the chips act , with the $100 billion and the accompanying investments of $220 billion so far that's been announced you put all of that together and you expect all of this to come online , and when everything that is now being promised is online , that number of 10% will

have risen to maybe about 29, . 30% at most .

You know we can , it's just manufacturing , it's the three Ds . Right , there's this commodity that we really don't have to worry . But what we actually found out we really don't have to worry , but what we actually found out was that there's actually a great deal of knowledge to be gained by having it together .

Yes , right , that engineers that are designing things can tweak things when they , when they are able to actually see it physically in front of them .

And by removing the manufacturing process , we actually became less competitive , not more competitive , and so it's a kind of a long winded back thing to you , john , of how will we , will we be able to , and how will we be able to kind of re-glue those things together .

And so I think the realization that even for areas like central Ohio that aren't connected necessarily to this type of work and placement , we felt the direct impact of that shortage and understood why there was this desire to bring in reshore , retool , retrain . But we had to kind of gear ourselves up to be ready for that .

You know it was once it was announced that Intel was looking to locate here in the central Ohio area .

It literally was an announcement in the next day , getting to work into business to look at this design , understand what it was that they do , how they do it , what it meant to us in terms of our foundation knowledge of the engineering process , how to adapt our student profile and our curriculum to what they were looking for .

It's been interesting because this core work with electrical , mechanical engineering that is somewhat transferable across the lines and that's what they really basically told us is that's the core of what we need to have as skill sets . But there is some specificity to the work of the semiconductor network that we need you to help students become more versed in .

They need to be understanding of clean room technology , of vacuum systems , because you all know that you have to have the utmost clean environment to work in , because even the smallest particles of dust on chips can be disastrous . So making sure that our students understood what that meant , how to work in that environment , what semiconductor processes really were .

Interestingly , you will appreciate this , our first cohort of students are going through this system currently . My youngest son is actually a student of ours in the Semiconductor Network and they are presenting their first Semiconductor 101 class presentations of their end result of knowledge and work . That's actually happening next week .

So we've been asked to four times our production of engineers in the component . Our institution , specifically the two-year college network in the state of Ohio , has been asked to create a 10 times network over the next 10 years .

The good news is there's a lot of interest in this . There's got to be a lot of education that goes with if people are interested in what this is . But how do I achieve it ? What kind of credentialing do I need to achieve ? What's the pay rate of return on that process ? It's an education system almost from ground zero and it's happening .

It is just it needs to happen so very quickly that we're putting a lot of resource generation into it but seeing a lot of results as well , and what is the pay rate ?

They're going to walk out with either a one-year certificate in this industry or two-year associates as that entry portal at about $62,000 . And in our network that translates fairly well .

Now imagine the semiconductor industry going through the build cycle that it now is , and just put yourself out at 2034 , 10 years from today , when all of these wafer fabs in all different parts of the countries are up and running and your job , your skill set , is still as valuable as it was the day you got certified and you now have 10 years of experience

under your belt . If you're willing to move to where the jobs are , you could be commanding a salary that's two to three times what it was when you first started . So you could be looking at $200,000 , depending on where you're willing to move to , which I think is a phenomenal change . Yeah .

Most people don't realize that Columbus is the 14th largest city in the country , so when we're talking about bringing people into the workforce , we were already seeing that as a growth arc . Intel has really highlighted the semiconductor network and even greater viability and we do see many , many individuals coming in specifically looking for that .

One of the other areas of great growth right now is what we would call the first generation American populations individuals , particularly from refugee , asylee capacities that have found Central Ohio as a secondary home .

We're doing a lot of training and development with that population art right now and , once we overcome some of the English language skills because that is a critical foundation the ESL components , they make a wonderful workforce addition stream to what we're trying to generate from adult learners that are coming into the areas and homegrown for our K through 12 programs

as well .

So that said that if you invested in Ohio , they wanted that the labor force be from Ohio and not be imported from elsewhere . So there was no precognition of where talent was . On the contrary , it was almost like let the chips fall where they may and then we'll figure out the supply chain that needs to go hand in hand with it from a skills perspective .

So much so that , to John's point , you know Central Ohio has never had a wafer fabrication facility . It's never been a found relocation . But they do have a rich history of being invested deeply in chemical technologies and in manufacturing . So that's a great combination .

So this chemical manufacturing process , the sustainability , the impact that may have environmentally , that was a great concern from some of the general populace and something we've had to work in tandem with Intel to make sure that we understand how to respond to that , and they have done a lot of public forum in relationship to that to lay and allay some of those

concern areas .

You contrast that with California . I mean I can't imagine anyone wanting to build mega fabs in California , given the kind of rules and regulations that we have with environmental factors in mind . So I think it's a choice that states are making , saying we're willing to , because it will fill in the blank , whatever the blank is in case of each of the states .

What we're going to have in California is a world where , yes , the chip designer makes a lot of money and the people who invest in those companies make a lot of money , but there aren't a lot of jobs for regular people . And I think that's part of the appeal of the Chips Act and what you're doing and what you're doing .

I guess the big question is are the federal requirements so pervasive and perverted that it makes it harder to do it ? I mean , you seem to feel like in Arizona that might be a problem .

A little bit of local knowledge perhaps in that particular regard , but again , it's years and years , it's decades , since they built a net new wafer fab there and they were up in arms recently in the face of the ship sack saying look , you know , insisting upon this being local doesn't make sense . There simply aren't people enough to build this fab .

It's just impossible for us to build this thing because we don't have the number of people who are qualified to build this fab . I mean , john referred to vacuum facilities and you know extreme clean rooms and things like that . Talking about it conceptually is one thing . Building something like that is really really hard .

It takes specialized skills and knowledge to do that . Where is that knowledge ? Well , it is where it chose to be , because 30 years ago was the last time I expressed that knowledge . Now I might have chosen to go settle in Alaska .

You can't say , well , you can't have an Alaskan working on an Arizonan facility , but he's the only guy who knows what to do about it . She's the only one who knows how to place this equipment in the right place .

That is not the case any longer , particularly with this type of industry sector . We are sharing with people in Kentucky , West Virginia , Michigan that these opportunities are here and if you want to be a part of that , we are welcoming you into our communities and we will help educate you in this design .

You know of people who are just relocating for what is obviously going to be a pretty sizable opportunity . If you're talking tens of thousands of people in central Ohio and a similar number in in in Arizona , and we have other fabs that are legacy fabs . That may like the Mid-Hudson Valley where I grew up near the Fishkill plant for IBM .

It'll be interesting to see whether these places get revitalized or not as a result .

And when once you stand up a fab , its lifespan is anywhere from eight to 15 years , depending on how much more investments go in . If you invest only once in eight years time it becomes sort of trailing edge technology , just naturally . But if you continue to invest it can last for 15 , 20 years . Some of them last even longer .

What they do is , if they have multiple wafer fabs , one gets mothballed but it becomes refurbished completely to be at the next leading edge and the next one , and so on and so forth . So Ohio I think the facility there is going to have 10 wafer fabs if it's fully constructed . If that be the case , they can play this catch-up game forever .

And around Nagoya , right around Nagoya , exactly right , does semiconductors work the same way ? Do you have supply regional suppliers , that kind of feed off of the main fab , or is it a different kind of a supply chain ?

So in large measure the equipment is made in California California , not in the United States . Made in California , made in Germany , made in the Netherlands and made in Japan . And these are different elements and they're almost mutually exclusive , and yet they're all an integral part of the manufacturing line that needs to come together .

Now that stock feed perhaps has the potential to be created into a Koretsu locally and there I think , when I think of Ohio , there is a natural advantage because they have been in chemicals for such a long time , they could repurpose themselves and say based on this , I could turn left and suddenly become a critical component supplier here .

But the stock portion of it is what you're saying is a relatively small part of the overall process , yes , but it has longer legs and then , whereas the equipment is one and done , the stock is forever . If you create a forever kind of a system there , John , are you seeing these ?

But Intel was very open with us about smaller supplier networks that they will need to have either groups locate in to Ohio or utilize pre-existing Ohio-based companies to provide some of that supplier engagement , not necessarily the large-scale fabrication mechanisms but all those supply .

And here's just a silly example in Pataskalo , ohio , which is about 10 miles from the fabrication center , we now have clean room suit producers and clean room suit cleaners , which we've never had before and never thought about but because of those bunny suits needed to be so viable in this clean room technology .

So it's those kind of smaller base supplier connections that we're seeing more and more of . But this system , as most of you all know , is a fairly secretive process . They keep a lot of information close to their vest and we are seeing supplier connectors come into our systems fairly rapidly , but we don't know them ahead of time . Oftentimes it's when they arrive .

Then we learn a little bit more about them and who they are , what they do and what can we do to help combine those networks .

Are you finding anything in the CHIPS Act that is politically difficult for you ? Yeah , you know , ohio is a red state basically , but there are some people on the right , for instance , who are against anything like the CHIPS Act and , of course , if you read the Wall Street Journal , they take a shot at it as much as they can .

How's the politics played out in Ohio ?

I would say that , and particularly in an election year , this becomes a fairly volatile issue given the fact that as a state we appreciate the economic impact of this , what it means in terms of job , workforce creation and even , candidly , notoriety of the state . You know , ohio doesn't always get the notoriety of a California , especially in this kind of space .

So there is that margin and we have a number of what I would call my diehard Republican friends and colleagues within our state legislature who do see that first and foremost before party loyalty .

But I can also tell you that it's a mixed bag and there are some on that side who really see this as a leftist type of organization and a process point that they don't necessarily support , more from the philosophical than the pragmatic .

So we're seeing some of that struggle , but at least from our governor's perspective , both the governor and lieutenant governor are profoundly entrenched on the platform of economic development and workforce enhancement .

So I would say to you we're in this for the long haul , regardless of that and , joel , you'll appreciate this , we're a red state until it becomes issue based , and then we're very , very purple and so I've got it .

So , if you step back and you think about the impetus for the CHIPS Act , the notion was to try to restore some power to the US , who arguably goofed up by having exported this technology out to begin with . Is it going to do that ?

This is the big final question Is the US really going to be able to create preeminence , or at least the alliance that they put together create preeminence that will make it strategically advantaged . What do you guys think about that ?

We didn't recognize , as early as our geopolitical opponents did , that the promise of artificial intelligence particularly comes with a lot of perils that are not yet understood , leave alone completely articulated and accounted for an open market focus , whereas our geopolitical opponents went at it from a state focus , very different perspectives .

By the time we woke up to the fact that this could spell disastrous results for nation states , including our own , I think the horse was out of the barn , so to speak . But we took rearguard action very fast , saying we will have export controls that will prevent leading-edge technologies from reaching the hands of our opponents on the global geopolitical stage .

And if you go back in history from a Taiwan perspective , being very specific to Taiwan , we've had lose support considerations for Taiwan , saying that they will not be attacked but there isn't a treaty .

They're not a treaty country with us , so to speak , but we've had joint exercises with them forever , almost 50 , 60 years , in support of there being a standalone , the fact that TSMC is there and is now manufacturing the most leading-edge technology chips in semiconductors is an obvious geopolitical threat .

Is an obvious geopolitical threat given the state of world affairs . So now the question becomes can you have just regulatory conformity preventing whatever technologies from reaching the hands of untoward opponents of untoward opponents ? That's an if on a good day , because the buyer-seller of any given transaction may only be visible up to a certain point .

There are after markets and after after markets , and so on and so forth . It's recursive to the end .

You know , there may have been a point in time I think there was a unipolar moment that lasted for about 10 years when the US said that these are the rules and the rest of the world said , yes , but those days are behind us , they're not ahead of us , which says that there are plenty of countries that can break rank and say well , we choose not to play by

your rules because they were your rules , not our rules . So that says that these market conditions that we're trying to now enact legislation with will rely upon cooperation from a large number of countries , if not all countries . When the threat was much more visible , like it was with nuclear fusion , I think it was easier .

But trying to figure out how to make leading-edge technology like artificial intelligence become real is very hard , because most people barely understand what people are talking about when they say artificial intelligence . It sort of reminds me of Daniel Kahneman , who recently passed away , as you know .

He was asked the question what do you think of artificial intelligence ? He said I'm too busy dealing with natural stupidity .

As we talked about earlier , we felt the chip manufacturing decline over COVID , with the supply chain , when you couldn't buy a new car or a truck because it wouldn't start because of the lack of chips accessibility . And it's funny because I appreciate what you were just saying too about nuclear warheads .

We understand , but I think we've all seen Terminator , that movie enough to know what really could happen if AI moves into a really negative state . So from our vantage point , we kind of see the role that we're playing as our localized addition to a globalized need . So for us it is a rallying point .

In some ways it's more of an altruistic purpose that , hey , you want to be a part of this next generation of technology , design , production , reshoring , retooling , resecuring our borders . This is something that you can do through the state of Ohio to impact that process . So again , when we talk about the interest generation , to impact that process .

So again , when we talk about the interest generation , having people look at this as a viable opportunity it is there . We're fielding literally hundreds of calls a week about what could this mean to me ? How can I engage in this ?

What would I be responsible for If we can couple the construction and the design of these fabrication centers to the nth degree that we've talked about ? Two has been a heavy lift , ten would be an enormous one .

But if we get the first two right and we do what we're duty bound to do for both construction and the workforce flow , I can see this being something that the state of Ohio has committed to for 50 plus years .