I had a kind of mixed reaction to this. First, love that they're taking questions. Loved the defense of American workers, in lockstep agreement there. Now for my disappointments: they spent over a year to decide to give money to TSMC, Intel, and Samsung. It's not a little curious they took forever to subsidize the status quo? I would have loved to see some follow-up on the notion they didn't want to consider any options that would not have happened without government money—wasn't the whole point to make leading edge happen in the US in a durable way because it wouldn't have happened without government money? Were they just hoping Intel would execute well? There's something about the celebratory attitude and defense of how great the team is that strikes me poorly, when just handing money to industry leaders to fix a problem doesn't feel like it required all that time, energy, and expertise.

Expand full comment

I understand the reaction, though I think the tough reality of the space is it takes a massive amount of expertise and massive amount of capital. That’s pretty much what makes semiconductor fabs so expensive and hard to make/maintain—enough so that there’s more or less just one firm left that’s truly on the bleeding, leading edge. (Maybe Intel’s 18A will turn that around!)

Based on reading this, which was helpful as well, this seems to be more to help barely unprofitable profits pencil out.

“So, we don't take a project that, you know, makes no sense. And we're like, great; we'll fund this too, you know, some absurd amount and make you build here, even if there's no economic rationale for it. This immediately means that we get good-caliber projects because these firms realize they're on the hook for the majority of the capital. And, so we see ourselves as taking a project that's, oh, it almost nets out, or maybe a smaller version of this project, or a less ambitious, ambitious version of this project nets out, or perhaps we do it on a slower timeframe. And we can enable it to happen faster, more ambitiously, and in a more economically sound way because the government is leaning in. I think that's one big one, we're not trying to create something that would not happen otherwise, and like an unnatural way.”

Unfortunately, most startups aren’t able to deliver these kinds of projects in scale or expertise. Nor can they bring the “majority of the capital.” In reality, all startups in semis are fabless these days, so production-wise, funding them doesn’t solve any issues.

I do think it’d be great to have programs that also fund more innovative, small companies as well, though that doesn’t seem to be the goal here.

Expand full comment

Kudos on getting an interview with these folks, and getting us a peek behind the curtain. Glad to hear that they’re thinking more deeply about this than simply splashing money around US semiconductors and the program office recruited smart people to run it.

Expand full comment