The US CHIPS Act provides about $53 billion in subsidies to bring semiconductor manufacturing back from Taiwan and South Korea, aiming for 0% to 20% of leading-edge chips made in the US by 2030. Todd Fisher, who runs the program, says the timing is right because the industry is in a downturn (high inventory, loose supply), making it ideal to 'fix the roof while the sun is shining.' AI demand has also boosted the need for advanced chips. Key holdings: TSMC (got subsidies, plans 3 US fabs, 14 suppliers follow), Intel (got subsidies, over $100 billion investment), Samsung (got subsidies, expanding to 2 fabs plus packaging, 24 suppliers tagging along).
This episode of Business Breakdowns examines the core issues of the 2022 U.S. CHIPS Act, which aims to enhance U.S. semiconductor manufacturing capabilities and reduce reliance on East Asia by providing nearly $53 billion in subsidies. This initiative responds to earlier chip shortages caused by dep
Guest Todd Fisher is the CIO of the CHIPS Act Office and a former senior investor at KKR. This episode focuses on how the U.S. CHIPS Act, with approximately $53 billion in subsidies, aims to reverse the offshoring trend of semiconductor manufacturing, as well as the execution progress and challenges of the program in 2024. Todd Fisher argues that the core goal of the CHIPS Act is not to create incremental capacity, but to "relocate" capacity that would have been built in Taiwan or South Korea to the United States—"This is not about adding X new fabs; rather, out of the original X fabs, 3 are now being built in the U.S."
Todd Fisher points out that the $39 billion in direct manufacturing incentives under the CHIPS Act is "not that large" relative to industry capital expenditure, but achieves a significant amplification effect through a carefully designed leverage structure.
> Fisher's own words: "We are 10 percent of what I just said of the capital that's investing. So that whole private equity thing, but alignment of interest, everybody's got skin in the game." Meaning, "We only account for 10% of the invested capital — that's alignment of interest: everyone has real money at stake."
Fisher emphasized that the CHIPS Act Office had defined its strategic objectives as early as early 2023, rather than passively responding to applications—this is a key distinction between this government program and traditional models.
The Four Pillars (Vision for Success, released in 2023):
1. Establish at least two leading-edge logic clusters in the United States
2. Significantly enhance advanced packaging capabilities
3. Build a commercially competitive leading-edge memory business
4. Address key bottlenecks in current mature process nodes (a fifth pillar was later added: building a supply chain to support all of the above)
Key Progress:
AI's Catalytic Role: Fisher acknowledged that "when the CHIPS Act was passed, AGI was not even a commonly used term," but the explosion in AI-driven demand has sharply increased the importance of leading-edge processes, making commitments such as TSMC's third fab and Intel's over $100 billion investment feasible.
Fisher emphasizes that the CHIPS Act office was built from scratch — when he joined in September 2022, there were "no people, no processes, no rules."
Workforce Challenge: Fisher believes the workforce is "one of the most critical determinants of success over the next decade" and that there is no silver bullet — "it must be advanced at the local level, every day, with a great deal of detail." The CHIPS Act office has a dedicated workforce team and has allocated specific funds for workforce ecosystem development in multiple announcements.
Fisher has made it clear that his primary objective is to "build America's economy and national security," rather than rewarding U.S. companies versus foreign companies.
Concerns about Chinese capacity: Fisher proactively mentioned that he is monitoring China's capacity expansion in mature process nodes, believing there is a risk of oversupply at certain nodes, which is part of his analytical framework.
Fisher emphasizes that funds are not disbursed in a lump sum on the announcement day but are released gradually through milestones.
Regarding election impact: Fisher states it is "basically negligible" — the CHIPS Act is "one of the most bipartisan projects of this administration," and bipartisan support has actually strengthened with the rise of China-related issues and AI. He personally focuses on "completing the award phase as quickly as possible and preparing the office for long-term operations."
| Position | Analyst View | Key Data |
|---|---|---|
| TSMC | Bullish (received incentives and expanded commitments) | 3 fabs; at least 14 suppliers plan to follow |
| Intel | Bullish (received incentives) | Over $100 billion in investment commitments |
| Samsung | Bullish (received incentives and expanded commitments) | 2 fabs + advanced packaging + R&D + expansion of Austin facility; approximately 24 strategic suppliers following |
| Micron | Bullish (received incentives) | Building leading high-bandwidth memory capacity in the U.S. (currently 0%) |
| SK Hynix | Slightly Positive | Announced construction of an advanced packaging facility in Indiana |
| GlobalFoundries | Positive (received incentives) | Secured a major long-term agreement with General Motors |
| Microchip | Positive (U.S.-based company) | No specific data provided |
1. Fisher: "This is not new capacity, it's relocated capacity" — TSMC originally planned to build X fabs in Taiwan, but now three of them are being built in the U.S. The core of the CHIPS Act is not about creating oversupply, but about changing geographic distribution.
2. $39 billion vs $650 billion — Direct incentives account for only about 6% of U.S. semiconductor investment over the next decade, but they have leveraged over $300 billion in total investment at a 1:10 ratio. Fisher emphasizes that "alignment of interests" is the core mechanism to prevent market distortion.
3. Timing of "fixing the roof while the sun shines" — The industry is currently in a cyclical trough (high inventory, ample supply), making it the best window to build resilience. Fisher uses the analogy of insurance: do not replace the roof when the wind is blowing.
4. Target from 0% to 20% for leading-edge manufacturing — The U.S. currently accounts for 0% of leading-edge logic manufacturing (>90% is in Taiwan), with a target of 20% by 2030. SIA forecasts it could reach 28% by the early 2030s.
5. AI as an unforeseen catalyst — When the CHIPS Act was passed, AGI was not yet a common term, but the demand explosion driven by AI has sharply increased the importance of leading-edge manufacturing, also prompting expansion commitments such as TSMC's third fab.
6. "Let more flowers grow" rather than picking winners — Fisher explicitly states no distinction between U.S. and foreign companies, only looking at long-term commitments (supply chain, R&D, workforce investment). The practical constraint is that certain technologies can currently only be done by specific companies.
7. Workforce is "the most critical success factor" with no silver bullet — After focusing on the workforce for four years at KKR, Fisher concludes that "it must be advanced at the local level, every day, with a great deal of detail." The CHIPS Act office has allocated dedicated funding for workforce ecosystem development.
8. Success metric is a "V-shaped curve" 5-10 years out — The U.S. currently accounts for about 10% of global semiconductor manufacturing capacity (which would drop to 8% without intervention), and SIA forecasts it could rebound to 14%. Fisher hopes to see a clear "V"-shaped reversal by 2030.