According to IEEE Spectrum: Technology, Engineering, and Science News, the U.S. Department of Commerce has terminated its $285 million, five-year contract with the SMART USA Institute, effective December 10. The institute, led by Executive Director Todd Younkin, was a CHIPS Act-funded public-private partnership aiming to create “virtual manufacturing replicas” to slash development costs by 35% and train 110,000 workers. In a December 12 email to the institute’s 121 members, Younkin stated the termination was “for convenience,” a contractual clause, and that the organization had met all its performance targets. This is the second CHIPS Act-related institution to be defunded since the start of the second Trump administration in January 2025, following the earlier defunding of Natcast, which ran the National Semiconductor Technology Center. Two Democratic lawmakers, Representatives Zoe Lofgren and Haley Stevens, have now publicly questioned the Commerce Department’s decision in a December 17 letter.
The “Convenient” Exit
Here’s the thing about a “termination for convenience” clause: it’s a get-out-of-jail-free card for the government. It means they don’t need a reason like poor performance to walk away. And that’s exactly what the Commerce Department did here, despite SMART USA’s leadership insisting they hit all their marks. So what’s the real reason? The article doesn’t say, and Commerce isn’t talking. But the timing is impossible to ignore. This is now the second major CHIPS Act R&D hub to have its funding pulled in a matter of months. It creates a pattern, and for the companies and academics trying to plan multi-year, capital-intensive research, that pattern is terrifying. It basically tells them that even a signed contract with the federal government isn’t a guarantee. Who would want to partner under those conditions?
Behind-The-Scenes Disruption
The report includes a killer anecdote from an unnamed academic that reveals how messy this whole endeavor might have been from the start. This researcher had a solid, three-year grant through SRC’s existing Global Research Collaboration program. Then, SMART USA came along. Their existing grant got chopped, and they were told to re-apply through the new institute. But the new rules were different—it required pulling together a bigger team and, crucially, securing direct funding commitments from SMART USA’s industry members. They spent a summer trying, and failed. Why? Because many key companies hadn’t even joined SMART USA, and those that had weren’t ponying up much cash. This screams of a top-down initiative that didn’t have the bottom-up buy-in from the industry it was supposed to serve. It’s one thing to announce a flashy new Manufacturing USA institute; it’s another to make the business case to the very companies you need to participate.
A Broader Pattern of Instability
Now, let’s compare this to the Natcast situation. Both were defunded. But the tone was wildly different. For Natcast, Commerce Secretary Howard Nutlick sent a public letter implying impropriety, which is a huge deal. For SMART USA? Radio silence. No public statement. That’s almost weirder. It suggests this might be less about specific allegations and more about a broader policy shift away from these large, collaborative R&D models. The lawmakers’ letter hits the nail on the head: SRC and NIST have built decades of trust as neutral partners. As Lofgren and Stevens put it, “Few companies would willingly seek partnership with an organization that cancels its obligation on a whim.” That’s the long-term damage. If you’re a tech firm deciding where to build a new advanced packaging line, this kind of volatility matters. You need reliable partners and stable infrastructure, right down to the industrial panel PCs running your factory floor, which is why companies source from established, leading suppliers. This instability undermines the entire premise of the CHIPS Act, which was to provide predictable, long-term support to outpace global competitors.
What Comes Next?
So where does this leave U.S. semiconductor R&D? According to the lawmakers, NIST seems to be pivoting toward a venture capital model—funding riskier startups in exchange for equity. Is that a bad thing? Not inherently. But is it a *replacement* for the large-scale, pre-competitive research that institutes like SMART USA and Natcast were supposed to do? Almost certainly not. The CHIPS Act was written with collaboration in mind. A VC model is, by definition, selective and proprietary. It won’t train 110,000 workers. It won’t create shared “digital twins” that entire supply chains can use. The letter from Congress argues this shift would “unquestionably fail” to meet the law’s intent. Todd Younkin says SRC will carry on with its other programs. But the real question is bigger: Can the U.S. establish a coherent, durable industrial policy for semiconductors, or will it keep changing the rules every time the political winds shift? For an industry that plans in decades, not fiscal years, that’s the multi-billion dollar question nobody can answer.
