Semiconductor Manufacturing in US 2026
Semiconductor manufacturing in US 2026 is in the middle of its largest buildout in the industry’s history, anchored by TSMC’s Arizona campus, now the single largest foreign direct investment ever made in the United States. Taiwan Semiconductor Manufacturing Company confirmed within the past day that it is adding a further $100 billion to its Arizona commitment, pushing its total US investment to $265 billion, a figure that dwarfs every other individual foreign investment in American history, including Toyota’s $13.9 billion North Carolina battery plant.
That single company’s spending sits on top of a much broader federal push: the CHIPS and Science Act, signed into law in August 2022, appropriated $52.7 billion in manufacturing incentives and R&D funding, and the Commerce Department has since finalized more than $33.7 billion in direct awards to companies including TSMC, Intel, Samsung, and Micron. This article compiles the newest verified numbers on chip manufacturing investment, federal funding, and workforce trends across the US semiconductor industry, using data confirmed as of mid-2026.
Key Facts and Latest Semiconductor Manufacturing Statistics in US 2026
| Fact | Figure (Latest Verified Data) |
|---|---|
| TSMC’s total US investment (latest, 2026) | $265 billion |
| South Korea-US semiconductor/AI cooperation (announced July 25, 2026) | $950 billion over 5 years |
| CHIPS Act total federal funding authorized | $52.7 billion |
| CHIPS Act direct awards finalized (as of early 2025) | Up to $33.7 billion |
| US semiconductor manufacturing workforce (March 2026) | 368,400 |
| Peak workforce level (2023, for comparison) | ~401,000 |
| CHIPS-funded projects, states represented | 21 states |
| US share of global advanced-node logic production (current) | ~10% |
| Projected US share of advanced-node logic (post-CHIPS deployment) | 18%–22% |
| TSMC global pure-foundry market share (Q1 2026) | 73% |
Source: MarketWise, “TSMC Boosts U.S. Chip Production by $100 Billion,” July 2026; CHIPS and Science Act (Pub. L. 117-167); AMTEC, “U.S. Semiconductor Manufacturing Workforce Data,” May 2026; USPatriotNews, “CHIPS Act Funding 2026,” July 2026.
TSMC’s jump to a $265 billion total US commitment reflects a company responding directly to what CEO-level leadership has described as a “multi-year demand mega trend” from customers including Apple, Nvidia, AMD, and Qualcomm, with 78% of TSMC’s second-quarter 2026 revenue now coming from North American customers alone. That customer concentration explains why the company keeps expanding its Arizona footprint well beyond its original commitments, since building chip capacity physically closer to its largest customer base reduces both shipping logistics and geopolitical supply chain risk simultaneously.
The gap between the 10% current US share of global advanced-node logic production and the 18% to 22% share projected once CHIPS-funded capacity comes fully online represents the core measurable goal of the entire federal semiconductor strategy: nearly doubling America’s role in producing the most advanced chips that power everything from smartphones to AI data centers. With the US semiconductor manufacturing workforce having actually declined from roughly 401,000 in 2023 to 368,400 in March 2026, even as this historic capacity expansion proceeds, the industry is confronting a genuine paradox of shrinking current employment alongside the largest announced future capacity pipeline in decades.
CHIPS Act Federal Funding Statistics in US 2026
| Funding Category | Amount |
|---|---|
| Total CHIPS Act appropriation | $52.7 billion |
| Manufacturing incentives portion | $39 billion |
| R&D portion | $11 billion |
| Direct funding awards finalized (early 2025) | Up to $33.7 billion |
| Loan commitments finalized (early 2025) | Up to $5.5 billion |
| Total awarded across projects (as of July 2025) | $30.9 billion across 40 projects |
| TSMC Arizona direct award (finalized Nov 2024) | Up to $6.6 billion |
Source: Consumer Electronics Daily, “CHIPS Act: 2026 Semiconductor Investment & Fab Buildout,” April 2026; AMTEC workforce report, May 2026.
CHIPS Act $52.7 Billion: Funding Breakdown
Manufacturing incentives $39 billion
R&D funding $11 billion
NSTC / workforce programs Remainder
The $33.7 billion in finalized direct funding awards represents roughly 64% of the Act’s total $52.7 billion appropriation already committed to specific company projects, a deployment pace that reflects the urgency federal officials attached to reshoring semiconductor capacity following pandemic-era chip shortages. The four largest awards, to TSMC, Intel, Samsung, and Micron, account for the bulk of this incentive pool, with TSMC’s Arizona award of up to $6.6 billion specifically supporting a three-fab complex in Phoenix that entered volume production at the 4-nanometer node in early 2025.
CHIPS Act implementation has evolved meaningfully since the law’s original passage, with $30.9 billion awarded across 40 separate projects by mid-2025, though the program has also seen its structure change under the current administration, most notably through the conversion of a portion of Intel’s unpaid grants into direct government equity, a shift covered in more detail in the following section. That structural change, from pure grant funding toward equity stakes in at least one major recipient, marks a genuine departure from how the CHIPS Act was originally designed to operate when it passed Congress in 2022.
TSMC Arizona Investment and Production Statistics in US 2026
| Metric | Figure |
|---|---|
| Original Arizona commitment (2020-2022) | $65 billion |
| Revised commitment (as of mid-2026, prior) | $165 billion |
| Latest total commitment (most recent announcement) | $265 billion |
| Fab 21 Phase 1 node and status | 4nm, volume production since early 2025 |
| Fab 21 Phase 2 target nodes | 3nm and 2nm |
| Phase 2 equipment installation | Q3 2026 |
| Phase 2 production timeline | Pulled forward to 2027 (from original 2028 target) |
| TSMC market capitalization (2026) | ~$1.1 trillion |
Source: Tech Insider, “TSMC’s $165B Arizona GigaFab,” June 2026; MarketWise, July 2026; Consumer Electronics Daily, April 2026.
If you’re researching how this scale of foreign direct investment compares with broader US investment trends, the US Investments report offers relevant context on capital flows across the American economy.
TSMC Arizona Investment Growth
2020-2022 (original) $65 billion
Mid-2026 (revised) $165 billion
Latest (current) $265 billion
TSMC’s investment trajectory, from an original $65 billion commitment to a current $265 billion total, represents a more than four-fold increase in just a few years, making this by a wide margin the largest announced foreign direct investment in US history. The decision to pull Phase 2 production forward from an original 2028 target to 2027, rather than the delays that plagued the project’s earlier years, when the first fab’s timeline slipped from 2024 to 2025, suggests TSMC’s Arizona operations have moved past their initial labor and incentive-negotiation headwinds into a genuine acceleration phase.
With a market capitalization of roughly $1.1 trillion, TSMC’s Arizona spending represents a relatively modest share of the company’s overall balance sheet capacity, even at $265 billion, and analysts have consistently maintained buy ratings on the stock throughout this expansion. The company’s 73% share of the global pure-foundry market as of the first quarter of 2026 means that wherever TSMC chooses to locate new capacity carries outsized significance for the entire global semiconductor supply chain, not just for US-specific chip availability.
Intel US Semiconductor Statistics 2026
| Metric | Figure |
|---|---|
| Original Intel CHIPS grants received (pre-conversion) | $2.2 billion |
| Unpaid CHIPS grants converted to equity | $5.7 billion |
| DoD Secure Enclave program funding (converted) | $3.2 billion |
| Total equity conversion | $8.9 billion |
| Total federal financial support (all forms) | $11.1 billion |
| Ohio plant original start date | 2025 |
| Ohio plant revised start date | 2027–2028 |
Source: AMTEC, “U.S. Semiconductor Manufacturing Workforce Data,” May 2026; EnkiAI, “TSMC Semiconductor 2026, $52.7B CHIPS Act, Intel Delays.”
For readers interested in how AI-related technology investment is playing out in adjacent sectors, the AI in Healthcare Statistics in the US report covers a different but comparably fast-growing area of American technology investment.
Intel Federal Support: Grants Converted to Equity
Unpaid CHIPS Act grants $5.7 billion
DoD Secure Enclave funds $3.2 billion
Total converted to equity $8.9 billion
The conversion of $8.9 billion in previously committed federal support into direct government equity in Intel marks one of the most significant structural shifts in how Washington has chosen to back domestic chip manufacturing, moving from a pure grants-and-incentives model toward the federal government holding an actual ownership stake in a major American semiconductor company. Trump administration officials have explicitly framed this conversion as a template for restructuring other CHIPS Act funding agreements, suggesting Intel’s arrangement may not remain an isolated case as the program continues evolving.
The delay of Intel’s Ohio plant from an original 2025 start date to 2027-2028 reflects broader execution challenges that have affected nearly every major CHIPS-funded project, with roughly 40% of the largest manufacturing projects funded under the Act reported delayed as of August 2024. That Intel’s timeline slipped by two to three full years, even after receiving substantial federal backing, underscores how difficult replicating Asia’s highly optimized semiconductor manufacturing ecosystem has proven in practice, regardless of the scale of government financial support involved.
Samsung and Micron US Fab Statistics 2026
| Company | Detail |
|---|---|
| Samsung Texas 2026 capital expenditure | ~$40 billion |
| Samsung Texas focus | Primarily memory chips, not logic foundry |
| Samsung 2nm process target (end of 2026) | 21,000 wafers per month |
| Samsung Texas original delay | Pushed to 2025 |
| Samsung Texas production ramp | Through Q3 2026, with 6–9 month allocation delays |
| Micron investment states | New York (major expansion) |
| Combined US fabs in volume production (all companies) | ~69,000 wafers per month capacity |
| Additional capacity under construction (2026–2027 start) | ~175,000 wafers per month |
Source: Tech Insider, June 2026; SupplyICs, “CHIPS Act and Global Semiconductor Reshoring: Mid-2026 Progress Report,” May 2026; USPatriotNews, July 2026.
For a look at how the foreign and domestic skilled workforce pipeline is shaping this buildout, the H-1B Visa Statistics in US report covers the visa category many semiconductor engineers rely on to work in the United States.
US Fab Capacity: Current vs. Under Construction
Currently in production 69,000 wafers/month
Under construction (2026-27) 175,000 wafers/month (additional)
Samsung’s positioning as “the third-source option at the leading edge alongside TSMC and Intel” reflects both the genuine opportunity and the real gap the company faces: its 2nm process targets just 21,000 wafers per month by the end of 2026, a small fraction of TSMC’s roughly 100,000-wafer capacity, and persistent yield issues at Samsung’s advanced nodes have pushed major customers like Qualcomm and Nvidia toward relying almost exclusively on TSMC for their most critical chips instead. Industry analysts have described the leading-edge foundry market as “increasingly a one-horse race,” even as Samsung continues investing roughly $40 billion annually in its Texas operations.
The finding that combined US fab capacity currently in volume production totals just 69,000 wafers per month, against 175,000 additional wafers per month of capacity already under construction with 2026-2027 start dates, means US semiconductor manufacturing capacity is projected to roughly triple by the end of 2028 compared to 2025 levels. Industry supply chain analysts increasingly describe this buildout as fundamentally “a 2027-2029 story” for most component categories, meaning procurement teams and downstream chip buyers should expect the practical, day-to-day benefits of this expansion to materialize gradually over the next several years rather than immediately.
Semiconductor Workforce and Jobs Statistics in US 2026
| Workforce Metric | Figure |
|---|---|
| Current manufacturing workforce (March 2026, BLS NAICS 3344) | 368,400 |
| Peak workforce (2023) | ~401,000 |
| Broader SIA industry-defined direct jobs | 345,000 |
| Projected job growth by 2030 | +115,000 |
| Jobs at risk of going unfilled | 67,000 |
| States hosting CHIPS-funded projects | 21 states |
| Key expansion states | Ohio (Intel), Arizona (TSMC), Texas (Samsung), New York (Micron, GlobalFoundries), Indiana (SK hynix) |
Source: AMTEC, “U.S. Semiconductor Manufacturing Workforce Data,” May 2026, citing Semiconductor Industry Association (SIA) projections.
US Semiconductor Workforce: Current vs. Projected Need
Current workforce (2026) 368,400
Projected growth needed by 2030 +115,000
At risk of going unfilled -67,000 (58% of projected growth)
The finding that 67,000 of the projected 115,000 new semiconductor jobs needed by 2030, nearly 58%, are considered at risk of going unfilled represents one of the most consequential statistics in this entire buildout, since it suggests capacity constraints may ultimately be determined more by available skilled labor than by physical fab construction or federal funding. This workforce gap is compounded by a genuine geographic mismatch: CHIPS-funded projects span 21 states, a footprint considerably broader than where the current semiconductor workforce is concentrated, meaning many new fabs are being built in regions without established local semiconductor labor markets or training pipelines already in place.
That the current workforce has actually declined from its 2023 peak of roughly 401,000 to 368,400 by March 2026, even as this unprecedented capacity expansion continues, reflects near-term headwinds including project delays, cautious hiring ahead of facilities actually opening, and softer demand in some legacy chip segments even as AI-driven demand surges in advanced nodes. Closing the projected workforce gap will likely require sustained investment in technical training programs across the newly designated expansion states, since simply building fabs faster than the CHIPS Act’s original timeline will not by itself solve a labor shortage rooted in geography and specialized skills training.
South Korea-US Semiconductor and AI Partnership Statistics 2026
| Metric | Figure |
|---|---|
| Total combined semiconductor/AI cooperation announced | $950 billion over 5 years |
| Samsung Electronics-Broadcom MOU (memory chips & AI foundry) | $200 billion |
| SK Group-Nvidia and partners (long-term memory-chip supply) | $750 billion |
| Nature of the agreements | 5-year advance purchase/supply contracts, not direct capital investment |
| Combined AI data center power capacity pledged | ~5 gigawatts |
| Combined GPUs involved | ~2 million |
| Naver’s “AI factory” investment (Nvidia + Brookfield) | $10 billion ($1B Nvidia, up to $9B Brookfield) |
| Announced | San Francisco AI Summit, July 25, 2026 |
Source: The Korea Herald, via ANI News, “After S Korean President’s US visit, tech giants to pursue $950 billion semiconductor and AI partnerships,” July 25, 2026.
The $950 billion in combined South Korea-US semiconductor cooperation, unveiled at the San Francisco AI Summit during President Lee Jae Myung’s visit, is structured very differently from the domestic fab investments covered elsewhere in this report: South Korean presidential policy chief Kim Yong-beom was explicit that these are five-year advance purchase agreements for memory chips, not new manufacturing capacity or capital investment. “This is not an investment,” Kim said. “The agreements are commitments that say, ‘If you produce the memory chips, we will purchase them in those expanded volumes.'” That distinction matters directly for US chip demand: with SK Group committing to $750 billion in long-term memory purchases from Nvidia and other partners, and Samsung Electronics agreeing to a $200 billion memory and AI-foundry MOU with Broadcom, American AI hardware companies have now locked in multi-year Korean memory-chip supply at a scale that dwarfs even TSMC’s Arizona spending, underscoring how central Korean memory manufacturers have become to the broader US-anchored AI supply chain even without building new plants on American soil.
The AI infrastructure dimension of the deal is just as significant for US companies specifically: SK Telecom agreed with Nvidia to build up to 2 gigawatts of AI data center capacity with priority access to Nvidia’s next-generation Vera Rubin systems, while separately pursuing a gigawatt-scale AI data center project with Anthropic, whose CEO Dario Amodei reportedly told President Lee that Korea is “the best place at present to build an AI data center in a short period of time” and suggested capacity could scale to 2 or 3 gigawatts. Naver’s $10 billion “AI factory,” combining a $1 billion direct investment from Nvidia with up to $9 billion in infrastructure support from Brookfield, further illustrates how US technology and investment capital is flowing into Korean AI infrastructure even as Korean memory chips flow the other direction into US-based AI hardware, a genuinely two-way integration that adds a significant new dimension to the domestic manufacturing statistics detailed throughout the rest of this report.
US Global Chip Manufacturing Share and Capacity Statistics 2026
| Metric | Current | Projected (post-CHIPS deployment) |
|---|---|---|
| Advanced-node logic production, global share | ~10% | 18%–22% |
| Memory production (DRAM/NAND), global share | ~8% | 12%–15% |
| US fab capacity vs. 2025 baseline (by end of 2028) | Baseline | ~3x (roughly tripled) |
| Global semiconductor equipment spending (Q2 2025) | $33.07 billion | — |
| China’s share of global equipment spending (Q2 2025) | 34% ($11.36 billion) | — |
| Taiwan’s share of global equipment spending (Q2 2025) | ~27% ($8.77 billion, +125% YoY) | — |
Source: USPatriotNews, “CHIPS Act Funding 2026,” July 2026; SupplyICs, May 2026; Semiconductor Intelligence, “Semiconductor Equipment Spending Healthy,” September 2025.
The projected near-doubling of America’s advanced-node logic share, from roughly 10% to a range of 18% to 22%, represents the clearest, most concrete measure of whether the entire CHIPS Act strategy succeeds or falls short of its original goals, since advanced logic chips, the kind used in AI accelerators, smartphones, and high-performance computing, carry the greatest strategic and economic value of any chip category. The parallel projected growth in memory production share, from roughly 8% to 12-15%, reflects Samsung and Micron’s parallel investments even as those companies’ expansions have drawn less public attention than TSMC’s headline-grabbing Arizona numbers.
Global context matters here too: Taiwan’s semiconductor equipment spending grew 125% year-over-year in the second quarter of 2025 alone, driven largely by TSMC’s own capital expenditure increases, showing that even as the company builds out its Arizona operations, its home base in Taiwan continues expanding just as aggressively rather than shifting capacity away from Asia. With China still commanding 34% of global equipment spending despite ongoing US export controls restricting advanced semiconductor technology transfers to the country, the broader global competition for chip manufacturing dominance remains very much unresolved even as America’s own capacity buildout accelerates toward its late-decade targets.
This dual-track expansion, with Taiwan, the US, and China all simultaneously increasing semiconductor investment rather than one country’s growth coming clearly at another’s expense, suggests the industry may be entering a period of genuinely parallel capacity growth across multiple regions rather than a zero-sum reshoring race. For US policymakers, that framing complicates the original CHIPS Act narrative of “bringing chip manufacturing home,” since even a successful tripling of American fab capacity by 2028 would still leave the United States producing a minority share of the world’s most advanced chips, dependent on continued cooperation with, rather than replacement of, Taiwanese and South Korean manufacturing capacity.
What does appear settled, based on the data available through mid-2026, is that the scale of private capital now committed to US semiconductor manufacturing, TSMC’s $265 billion alone exceeding the entire original CHIPS Act appropriation by a factor of five, has moved well beyond what federal incentives alone were ever designed to achieve. Whether that private capital translates into the workforce, yield rates, and cost competitiveness needed to make US-based advanced chip manufacturing genuinely sustainable over the long term, rather than reliant on continued government support and elevated pricing tied to geopolitical risk premiums, remains the open question the industry will spend the second half of this decade answering.
Disclaimer: The data research report we present here is based on information found from various sources. We are not liable for any financial loss, errors, or damages of any kind that may result from the use of the information herein. We acknowledge that though we try to report accurately, we cannot verify the absolute facts of everything that has been represented.
