(To receive weekly emails of conversations with the world’s top CEOs and business decisionmakers, click here.)

As part of an effort to regain its position as a leading maker of semiconductors amidst a global chip shortage, Intel confirmed to TIME this week that it will commit $20 billion to build a manufacturing mega-site in New Albany, on the outskirts of Columbus, Ohio. The chip maker said it will build at least two semiconductor fabrication plants, or fabs, on the 1,000-acre site, where Intel will research, develop, and manufacture its most cutting-edge computer chips, employing at least 3,000 people. Construction will begin this year and the plant should be operational by 2025, the company said.

The announcement came amidst a push to increase domestic manufacturing of semiconductors. Partly because of enormous incentives offered by other countries to jumpstart semiconductor manufacturing on their shores, the share of chips made in the U.S. has fallen to 12%, from 37% in 1990, according to the Semiconductor Industry Association(SIA). As booming demand and supply chain woes led to semiconductor shortages over the past year, entire U.S. industries like auto manufacturing were crippled.

To create a more reliable supply of chips, the federal government is weighing providing incentives for chip makers in the U.S. The CHIPS for America Act, passed last year, authorized federal investments in chip manufacturing, but it did not provide funding. The Senate passed $52 billion in funding in June, but the House has not passed the legislation.

Intel CEO Pat Gelsinger spoke with TIME about his plans to build "the largest silicon manufacturing location on the planet" in Ohio, bring the entire semiconductor supply chain stateside, and get congress to support it all.

This interview has been condensed and edited for clarity.

You've referred to yourself before as a Pennsylvania farm boy. Does your affinity for the Midwest play into this choice of Ohio for the new factories?

I will say that we looked at the talent. And Ohio has a great reputation as a manufacturing site. For other industries, it's been well leveraged in that regard—manufacturing is hard work. It is just very disciplined execution. We always sort of joked about in Pennsylvania, Ohio was the state next door, but the Heartland has a great track record building at scale manufacturing. I joked with the team that we helped to establish the Silicon Valley, you know, the Silicon Forest, the Silicon desert, the Silicon Oasis, and now we're going to do the Silicon Heartland.

I know some analysts have said—and you've said publicly—that Intel stumbled, but now you've taken the reins. How do you see this mega fab as part of your new vision for the company?

Clearly, hey, we have to fix some of the stumbles. We feel like we're well underway on putting our engine back in place and as I said on the last earnings call, Intel 7, 4, 3, 28, and 18A—five process nodes in four years—are all on or ahead of schedule. We need the recipe, if you would, to put in the factories but you also need the factories of scale. So we're going to build out an extraordinary manufacturing engine that can build leading technology with leading capacity. There's a winning combination.

Unfortunately, it takes several years to build factories. So I’ve just pushed the team super aggressively to get started. We want to set new records for how rapidly we're able to get the fabs built and operational. We are really looking forward to the warm reception that we've seen from Ohio to quickly get land established, to get environmental building permits, electricity, water, all these things that are required to get in place. They've just really demonstrated that they're committed to making this successful with us.

Was one of the reasons for choosing Ohio that it did seem like you'd be able to do this more quickly than some of the other locations?

We've just done enough of these now that things like environmental approvals, land access and being able to have the support of the local community—it's so easy to lose a quarter or two on a project like this. It's not like there’s ill will, but there's a lot of stuff to get through when you're building a site of this capacity.

You have big facilities in Oregon and Arizona. You could just double down on those, right? Why start from scratch, especially when some of Intel's problems in the past have been attributed to trying to do too much at once?

Part of it was that we are at considerable scale at these other locations now. We have $50-plus billion of capital in Oregon, and similarly in Arizona. We're somewhat tapping the limits of the sites as we have. That's a talent statement and that’s a construction statement. If I said, let's do four new fabs in Arizona, well, every concrete truck is already busy. That’s a part of it.

Part of it as well, as you know, this is a big site. We’re talking about a 1,000 acre site. We didn't have any adjacent parcels at those other locations that we could immediately scale into. We've also seen that it's been a long time since we started a major new site. It's been several decades since we've done that. And those were very successful cycles for the company that essentially afforded us first mover advantages. We were the first major company in Oregon, the first major company in Arizona, the first major company in Ireland and Israel, and beyond, and they've become talent magnets for us over a sustained period of time.

We think this is going to open up a period of time of extraordinary community goodwill. Hey, if you're an Ohio State Buckeye, you know where you want to come and work now. It really reaches into the whole Midwest engineering talent base. So we just saw a lot of these different pieces coming together very nicely to start this next site.

Some analysts questioned IDM 2.0 [Intel's strategy for expanded manufacturing capacity], saying, 'Who are you building the capacity for?' Do you see this as being more for the foundry side of things, or more for the Intel side of things? How does this new site play into IDM 2.0?

Well we've used the language of ‘SMART capital,’ which basically has two notions to it. SMART Capital says, basically, build shells. If you look at a $10 billion fab, you spend the first $2 billion, and then you spend the $8 billion; you spend a lot of the time building it and then you spend a lot of the money putting equipment into it. So the first thing is we want to get shells built so we have more flexibility. It is well-chronicled now that the industry is atrociously short of capacity. So we just need to get all these shells built so that we can populate it.

And then the second aspect of SMART capital is to leverage the foundry or our internal product needs. To some degree, I don't care what the answer to that question is, as long as we see that there are good uses for capacity on the horizon. The reality is we expect that there's going to be a good mix of foundry capacity, and our internal product capacity. But that's part of the IDM 2.0 strength is that we can start these long term four- and five-year projects, without necessarily being able to precisely answer how many wafers are for foundry or how many wafers are products.

I also expect the answer to that is going to be dynamic. This week, we're running more foundry capacity. Next week, we're running more internal products capacity. That's part of the strength of IDM 2.0 is that we get to sum the needs together for both of those areas of our business and get optimal capacity. And capital utilization through both.

In Arizona, you started building and then demand fell off, I believe in 2014. And you kind of put that aside for a little bit. Is that a concern here? That there's huge demand now, but maybe there isn't going to be the demand in the next couple of years?

We believe that we are in a 10 year cycle of strong demand. The factory that we will build is capable of our most advanced process technology. We see almost an insatiableness to leading edge capacity. Do you want your cell phone to be more capable and have better battery life? Do you want your AI to be more intelligent? Do you want your car to have more EV AV characteristics? Do you want your gaming PC? Do you want your AR VR? All of these are based on the most advanced capabilities. So on the one hand, we don't see an end in sight. At the same time, we always know there are economic cycles as well.

And that sort of brings us back to the SMART Capital. We're gonna build a site out rapidly and then we'll populate it with equipment, the more expensive equipment as the demand signals become more clear. But I think I'm catching up with demand for a decade. We're launching new products in very substantial markets. In addition to our core server and data center business, our autonomous vehicle business – the car industry is increasing their use of semiconductors by 5x over the decade – we're watching our graphics and high performance computing business, a market that we're zero in and that's a $20 billion market. So we're going to be expanding. We’re launching the foundry business as we've talked about, where today we're zero, and it's already a $150 billion industry.

You’ve talked about this mega fab before as a $60 to $120 billion investment. This announcement is much smaller. What's it going to take to get to that big number, or is that kind of off the table now?

The site is such that we'll be able to scale. The initial site is suitable for eight fabs, each fab is about $10 billion. So essentially, the full site project would be up to that 80 to 100 billion investment. We're starting with the first agreements with the city and Ohio and the region for the first two modules, which is the 20 billion—essentially exactly what we did in Arizona. We hope we're doing module 3, 4, 5, 6, as we build out over the decade. And in addition to that we also have some land next to that first eight pack that we could even grow larger over time or with other facilities there. That's why the language I'll be using is that our expectation is this becomes the largest silicon manufacturing location on the planet. Period. By us or anybody else in the industry.

What else would go there?

Offices, other suppliers would be building around it. But also in our long term plan, maybe we bring some package, assembly, test capabilities to the site as well. Some had said, hey, you know, bringing the fabs on US soil is super cool. But supply chain needs other areas of the package assembly test, right? My objective would be if we get the CHIPS Act done, that also accommodates building out package assembly test capabilities. While it's not as critical as the fabs, it's still very critical. And my objective would be sand to product to services, all on American soil; integrated supply chain capacity to meet our economic and national security requirements as well. So those types of facilities might be other things that we might be putting at the site as well, over time.

Chips cross borders so many times, but you see a day where it could potentially almost all be here?

That would be my objective. That we would have rebuilt a complete supply chain. I expect that we're still going to be doing the preponderance of our package assembly tests outside of the US, but for more critical products, particularly for those for Defense Intelligence, I think there's a strong desire to have that full supply chain being on American soil. And we want to be in a position that we've accommodated that.

As technologies keep moving forward, why did we end up in Asia? There's really two reasons driving it. There was the need for low cost labor on the lower end of the supply chain. And then the second was the consolidation of supply chains. The third was the very aggressive pursuit by those countries.

Now let's say in the second half of this decade, or going into next decade, we're doing a lot more automation, and low cost labor isn't as critical. With the CHIPS Act, we've leveled the playing field with the cost of capital and other infrastructure requirements. We've rebuilt below the critical supply chains on American soil. When we look to 2030 and beyond, we could be at a dramatically different view of our national supply chains. And, as we've said, COVID has induced us to think about these things differently.

The $52 billion proposed in the CHIPS Act is a lot of money on the one hand, but on the other hand it’s spread out among all the chip companies here. It doesn't seem like that much. Can you talk about how much of this plan needs the CHIPS Act?

We would say it's a force multiplier. Because for every dollar that is getting the capital offset, there's at least $3 if not $4 or $5 being invested. So $50 billion times four is $200 billion. Okay, now this is a pretty interesting number. And, you know, hey, I look to be a meaningful recipient of that. Do we also believe TSMC, Samsung, and others should be beneficiaries? Not as much as us. We think American IP should be benefited more significantly. Any US taxpayer would say the next Samsung fab on US soil is good, the next Intel fab on US soil is better.

We're not waiting for the CHIPS Act. I announced Arizona, and I said, please get the CHIPS Act. Now we're announcing Ohio, and I'm gonna say please, please get CHIPS Act done. It allows us to go bigger and faster, but we're not waiting on its passage to get underway. We're running a company. We're not running a political process. We're getting going.

Talk to me a little bit about that, because I think one of the interesting things about the CHIPS Act is that you're kind of having to ally with all your competitors to talk to Congress, at a time when nothing is getting done in Congress. How have you worked with other companies to get this moving?

Yeah, that's probably best embodied by the SIA Semiconductor Industry Association. It has the fabless guys—Qualcomm, AMD—as well as the fabfull guys like TI, Micron. We’re all together in the room. It's very bipartisan, bicameral, right. Ohio happens to have a Republican and a Democratic Senator. The CHIPS Act was led by Cornyn and Schumer, Republican and Democrat from Texas and New York. And I've met with the problem solvers caucus of congressional leaders, both Democrats and Republicans.

All in person?

Yes, in person. So our industry association is being active. We've also reached out to another number of groups like the Manufacturing Association. You know, the telecommunications industry associations came alongside and said fund this. I'm also spending significant time with defensive intelligence. My first meeting with the Undersecretary of Defense, I basically scolded her. I said, why am I explaining why this is so important to Congress, and you're not? It is a national priority.

Are your competitors alongside you in those meetings?

Some of them are. My head of government affairs says, 'We have to make sure you stay below the lobbyist threshold, Pat.'

I just want to talk to you a little bit about your vision for Ohio. What do you see the New Albany area looking like in five or 10 years?

We would say, if you go to the Chandler[, Ariz.] or the Oregon facility, we want to do that. What's happened at those locations, is we've seen the sites just grow over time. These are good jobs. Everything from technician jobs to engineering jobs, research PhD, but it's also the construction site, and then it becomes a magnet. We need all of our supply chain around us. We need all of the other companies that support the fab. So that becomes this concentric ring around it. And then if you build up all the infrastructure, other companies start to come around that as well. And it just gets bigger and bigger. If you go to the Hillsborough region up in Oregon, you do feel like you've shown up in a mini Silicon Valley. There's the Google research site and there's the Microsoft data center and over there's the Google super scale cloud center. It just becomes this magnet of innovation. It's what we saw in Israel when we first went there almost 50 years ago.

So, would you say New Albany will be unrecognizable in 10 years?

Well, how many people think about New Albany as the hub of the largest semiconductor manufacturing location on the planet? This is powerful, it’s transformative to the region. And yeah, I mean, little New Albany today versus the high tech mega manufacturing center of the heartland in five years? Yeah, it's gonna be unrecognizable.